Model profile · Zhipu

glm-5.3-max

Closed weightsmid · registry tag 2026 benchmark-derived

Thin verified coverage

Reads as thin verified coverage across the resolved source data. Capability fit excludes speed and cost.

Visible coverage: 45.9%
Verified coverage: 45.9%
Spread: 95.4%
Last verified: Oct 3, 2026

72%capability fit

textcodedocument5 aliases35 official source links

Open compare

Data version

Current snapshot.

Published Oct 3, 2026Model list checked4 stale sources105 claim warnings9 providers · 1334 tracked models

The registry snapshot and page stamp are shown so a stale deploy is visible at a glance. View live data status.

Source-linked scores by benchmark

Each row keeps the benchmark source, source type, raw metric, and percentile inside its fair comparison set. Capability summaries exclude archived categories, speed, and cost; those rows remain below.

Thin verified coverageThis model currently reads as thin verified coverage across the resolved source data.

Chat / text35 benchmarks81.4%

Intelligence Index

AA · Chat / text · Combined

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #34 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 45
Percentile: 94.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `intelligenceIndex`.

94.1% percentile inside its fair comparison set

45Raw benchmark value

AA-Omniscience accuracy

AA · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #109 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 33.9%
Percentile: 76.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `omniscienceAccuracy`.

76.2% percentile inside its fair comparison set

33.9%Raw benchmark value

AA-Omniscience non-hallucination

AA · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #29 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 70.4%
Percentile: 93.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `omniscienceNonHallucination`.

93.8% percentile inside its fair comparison set

70.4%Raw benchmark value

Input price

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #252 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $1.4 /1M input tokens
Percentile: 29.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mInputTokens`.

29.4% percentile inside its fair comparison set

$1.4 /1M input tokensRaw benchmark value

Output price

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #241 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $4.4 /1M output tokens
Percentile: 33.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mOutputTokens`.

33.3% percentile inside its fair comparison set

$4.4 /1M output tokensRaw benchmark value

Output Speed

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #162 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 70.7 tokens/s
Percentile: 38.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianOutputTokensPerSecond`.

38.5% percentile inside its fair comparison set

70.7 tokens/sRaw benchmark value

Time to first token

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #153 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 3.42s
Percentile: 42.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianTimeToFirstTokenSeconds`.

42.4% percentile inside its fair comparison set

3.42sRaw benchmark value

Time to first answer token

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #180 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 31.70s
Percentile: 31.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianTimeToFirstAnswerTokenSeconds`.

31.7% percentile inside its fair comparison set

31.70sRaw benchmark value

Text Arena

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #27 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,478
Percentile: 93.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: overall. Source rank: #27. Votes: 17857. Organization: zai. License: MIT.

93.1% percentile inside its fair comparison set

1,478Raw benchmark valueCI 1,473 - 1,484

Text Arena · Creative Writing

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #33 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,454
Percentile: 91.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: creative_writing. Source rank: #34. Votes: 4034. Organization: zai. License: MIT.

91.4% percentile inside its fair comparison set

1,454Raw benchmark valueCI 1,444 - 1,464

Text Arena · English

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #23 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,490
Percentile: 94.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: english. Source rank: #23. Votes: 7223. Organization: zai. License: MIT.

94.1% percentile inside its fair comparison set

1,490Raw benchmark valueCI 1,482 - 1,497

Text Arena · Exclude Ties

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #28 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,488
Percentile: 92.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: exclude_ties. Source rank: #28. Votes: 13198. Organization: zai. License: MIT.

92.8% percentile inside its fair comparison set

1,488Raw benchmark valueCI 1,481 - 1,495

Text Arena · Hard Prompts

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #25 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,503
Percentile: 93.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: hard_prompts. Source rank: #25. Votes: 11434. Organization: zai. License: MIT.

93.6% percentile inside its fair comparison set

1,503Raw benchmark valueCI 1,497 - 1,510

Text Arena · Hard Prompts English

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #18 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,512
Percentile: 95.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: hard_prompts_english. Source rank: #18. Votes: 4618. Organization: zai. License: MIT.

95.5% percentile inside its fair comparison set

1,512Raw benchmark valueCI 1,503 - 1,521

Text Arena · Instruction Following

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #25 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,476
Percentile: 93.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: instruction_following. Source rank: #25. Votes: 6049. Organization: zai. License: MIT.

93.6% percentile inside its fair comparison set

1,476Raw benchmark valueCI 1,467 - 1,484

Text Arena · Longer Query

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #33 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 91%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: longer_query. Source rank: #34. Votes: 8337. Organization: zai. License: MIT.

91% percentile inside its fair comparison set

1,486Raw benchmark valueCI 1,479 - 1,494

Text Arena · Multi Turn

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #39 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,480
Percentile: 89.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: multi_turn. Source rank: #41. Votes: 2694. Organization: zai. License: MIT.

89.8% percentile inside its fair comparison set

1,480Raw benchmark valueCI 1,468 - 1,492

Text Arena · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #25 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,471
Percentile: 93.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: overall. Source rank: #26. Votes: 17857. Organization: zai. License: MIT.

93.6% percentile inside its fair comparison set

1,471Raw benchmark valueCI 1,466 - 1,477

Text Arena · Creative Writing · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #23 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,457
Percentile: 94.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: creative_writing. Source rank: #24. Votes: 4034. Organization: zai. License: MIT.

94.1% percentile inside its fair comparison set

1,457Raw benchmark valueCI 1,447 - 1,467

Text Arena · English · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #19 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,482
Percentile: 95.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: english. Source rank: #19. Votes: 7223. Organization: zai. License: MIT.

95.2% percentile inside its fair comparison set

1,482Raw benchmark valueCI 1,474 - 1,490

Text Arena · Exclude Ties · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #26 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,477
Percentile: 93.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: exclude_ties. Source rank: #27. Votes: 13198. Organization: zai. License: MIT.

93.4% percentile inside its fair comparison set

1,477Raw benchmark valueCI 1,469 - 1,484

Text Arena · Hard Prompts · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #21 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,488
Percentile: 94.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: hard_prompts. Source rank: #21. Votes: 11434. Organization: zai. License: MIT.

94.7% percentile inside its fair comparison set

1,488Raw benchmark valueCI 1,482 - 1,495

Text Arena · Hard Prompts English · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #16 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,497
Percentile: 96%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: hard_prompts_english. Source rank: #16. Votes: 4618. Organization: zai. License: MIT.

96% percentile inside its fair comparison set

1,497Raw benchmark valueCI 1,488 - 1,506

Text Arena · Instruction Following · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #23 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,477
Percentile: 94.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: instruction_following. Source rank: #23. Votes: 6049. Organization: zai. License: MIT.

94.1% percentile inside its fair comparison set

1,477Raw benchmark valueCI 1,469 - 1,485

Text Arena · Longer Query · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #20 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,483
Percentile: 94.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: longer_query. Source rank: #20. Votes: 8337. Organization: zai. License: MIT.

94.6% percentile inside its fair comparison set

1,483Raw benchmark valueCI 1,476 - 1,490

Text Arena · Multi Turn · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #31 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,472
Percentile: 92%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: multi_turn. Source rank: #33. Votes: 2694. Organization: zai. License: MIT.

92% percentile inside its fair comparison set

1,472Raw benchmark valueCI 1,460 - 1,484

Instruction following

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #33 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 69.3%
Percentile: 50.8%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: IF. Tasks scored: 4.

50.8% percentile inside its fair comparison set

69.3%Raw benchmark value

Language

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #31 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 79.9%
Percentile: 53.8%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Language. Tasks scored: 3.

53.8% percentile inside its fair comparison set

79.9%Raw benchmark value

Paraphrase

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #40 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 64.7%
Percentile: 40%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: paraphrase. Category: IF.

40% percentile inside its fair comparison set

64.7%Raw benchmark value

Simplify

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #29 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 65.9%
Percentile: 56.9%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: simplify. Category: IF.

56.9% percentile inside its fair comparison set

65.9%Raw benchmark value

Story generation

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #31 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 72.1%
Percentile: 53.8%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: story_generation. Category: IF.

53.8% percentile inside its fair comparison set

72.1%Raw benchmark value

Summarize

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #22 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 74.6%
Percentile: 67.7%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: summarize. Category: IF.

67.7% percentile inside its fair comparison set

74.6%Raw benchmark value

Connections

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #45 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 95.3%
Percentile: 32.3%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: connections. Category: Language.

32.3% percentile inside its fair comparison set

95.3%Raw benchmark value

Plot unscrambling

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #37 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 60.2%
Percentile: 44.6%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: plot_unscrambling. Category: Language.

44.6% percentile inside its fair comparison set

60.2%Raw benchmark value

Typos

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #23 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 84%
Percentile: 80%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: typos. Category: Language.

80% percentile inside its fair comparison set

84%Raw benchmark value

Coding21 benchmarks73.6%

SciCode

AA · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #14 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 59%
Percentile: 93.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `scicode`.

93.4% percentile inside its fair comparison set

59%Raw benchmark value

Text Arena · Coding

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #34 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,521
Percentile: 91.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: coding. Source rank: #34. Votes: 4424. Organization: zai. License: MIT.

91.1% percentile inside its fair comparison set

1,521Raw benchmark valueCI 1,511 - 1,530

Text Arena · Coding · No Style Control

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #27 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,497
Percentile: 93%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: coding. Source rank: #27. Votes: 4424. Organization: zai. License: MIT.

93% percentile inside its fair comparison set

1,497Raw benchmark valueCI 1,487 - 1,506

Vibe Code Bench v1.1

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #23 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 78.1%
Percentile: 76.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: vibe-code; provider: Zhipu AI.

76.3% percentile inside its fair comparison set

78.1%Raw benchmark valueCI 70.4% - 85.8%

WebDev Arena

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #19 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,623
Percentile: 85.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: overall. Source rank: #20. Votes: 7658. Organization: zai. License: MIT.

85.8% percentile inside its fair comparison set

1,623Raw benchmark valueCI 1,615 - 1,631

Code Arena · Webdev Html

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #25 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,590
Percentile: 80.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: webdev-html. Source rank: #26. Votes: 1016. Organization: zai. License: MIT.

80.8% percentile inside its fair comparison set

1,590Raw benchmark valueCI 1,569 - 1,610

Code Arena · Webdev React

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #17 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,641
Percentile: 85.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: webdev-react. Source rank: #18. Votes: 5094. Organization: zai. License: MIT.

85.8% percentile inside its fair comparison set

1,641Raw benchmark valueCI 1,631 - 1,651

Code Migration

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #21 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 44.2%
Percentile: 70.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: code-migration; provider: Zhipu AI.

70.6% percentile inside its fair comparison set

44.2%Raw benchmark valueCI 35.8% - 52.6%

IOI

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #12 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 68.4%
Percentile: 70.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: ioi; provider: Zhipu AI.

70.3% percentile inside its fair comparison set

68.4%Raw benchmark valueCI 53.8% - 83.1%

ProgramBench

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #13 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 1.5%
Percentile: 81%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: programbench; provider: Zhipu AI.

81% percentile inside its fair comparison set

1.5%Raw benchmark valueCI 0% - 3.2%

Terminal-Bench 2.1

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #25 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 71.5%
Percentile: 64.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: terminal-bench-2-1; provider: Zhipu AI.

64.2% percentile inside its fair comparison set

71.5%Raw benchmark valueCI 67.7% - 75.4%

SWE-bench Verified

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #6 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 95.4%
Percentile: 95.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: swebench; provider: Zhipu AI.

95.1% percentile inside its fair comparison set

95.4%Raw benchmark valueCI 93.6% - 97.2%

LiveCodeBench

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #65 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 80.5%
Percentile: 48.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: lcb; provider: Zhipu AI.

48.8% percentile inside its fair comparison set

80.5%Raw benchmark valueCI 78.4% - 82.6%

Terminal-Bench 4.0

TERMINAL-BENCH · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #16 · Source label: GLM-5.3 (max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Terminal-Bench
Raw value: 41.8%
Percentile: 42.3%
Last updated: aging
Eligibility: headline eligible

Parsed from the public Terminal-Bench 4.0 official published leaderboard (status: display). Collapse policy: highest verified score per canonical model. Selected agent: Claude Code. Display model: GLM-5.3. Agent URL: https://claude.com/product/claude-code. Reported stderr: 1.648 percentage points.

42.3% percentile inside its fair comparison set

41.8%Raw benchmark valueCI 38.6% - 45.1%

Agentic coding

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #15 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 60.9%
Percentile: 78.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Agentic Coding. Tasks scored: 3.

78.5% percentile inside its fair comparison set

60.9%Raw benchmark value

Coding

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #26 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 79%
Percentile: 61.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Coding. Tasks scored: 2.

61.5% percentile inside its fair comparison set

79%Raw benchmark value

JavaScript

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #12 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 72.7%
Percentile: 87.7%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: javascript. Category: Agentic Coding.

87.7% percentile inside its fair comparison set

72.7%Raw benchmark value

TypeScript

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #17 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 50%
Percentile: 81.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: typescript. Category: Agentic Coding.

81.5% percentile inside its fair comparison set

50%Raw benchmark value

Python

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #21 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 60%
Percentile: 80%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: python. Category: Agentic Coding.

80% percentile inside its fair comparison set

60%Raw benchmark value

Coding generation

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #35 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 77.5%
Percentile: 52.3%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: code_generation. Category: Coding.

52.3% percentile inside its fair comparison set

77.5%Raw benchmark value

Coding completion

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #23 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 80.4%
Percentile: 78.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: code_completion. Category: Coding.

78.5% percentile inside its fair comparison set

80.4%Raw benchmark value

Reasoning / math / science18 benchmarks68.1%

Humanity's Last Exam

AA · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #53 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 42.3%
Percentile: 90.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `hle`.

90.2% percentile inside its fair comparison set

42.3%Raw benchmark value

GPQA

AA · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #35 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 91.7%
Percentile: 93.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `gpqa`.

93.1% percentile inside its fair comparison set

91.7%Raw benchmark value

CritPt

AA · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #47 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 19.1%
Percentile: 89.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `critpt`.

89.9% percentile inside its fair comparison set

19.1%Raw benchmark value

Text Arena · Math

AR · Reasoning / math / science · Human

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #20 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,494
Percentile: 94.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: math. Source rank: #20. Votes: 739. Organization: zai. License: MIT.

94.7% percentile inside its fair comparison set

1,494Raw benchmark valueCI 1,473 - 1,515

Text Arena · Math · No Style Control

AR · Reasoning / math / science · Human

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #19 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,491
Percentile: 95%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: math. Source rank: #19. Votes: 739. Organization: zai. License: MIT.

95% percentile inside its fair comparison set

1,491Raw benchmark valueCI 1,470 - 1,512

ProofBench

VALS-AI · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #26 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 49%
Percentile: 40.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: proof_bench; provider: Zhipu AI.

40.5% percentile inside its fair comparison set

49%Raw benchmark valueCI 39.2% - 58.8%

GPQA Diamond

VALS-AI · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #36 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 88.1%
Percentile: 72.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: gpqa; provider: Zhipu AI.

72.6% percentile inside its fair comparison set

88.1%Raw benchmark valueCI 84.5% - 91.8%

MMLU Pro

VALS-AI · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #38 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 86.8%
Percentile: 70.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: mmlu_pro; provider: Zhipu AI.

70.4% percentile inside its fair comparison set

86.8%Raw benchmark valueCI 86.1% - 87.4%

Mathematics

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #43 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 87.9%
Percentile: 35.4%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Mathematics. Tasks scored: 4.

35.4% percentile inside its fair comparison set

87.9%Raw benchmark value

Reasoning

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #35 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 85.8%
Percentile: 47.7%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Reasoning. Tasks scored: 4.

47.7% percentile inside its fair comparison set

85.8%Raw benchmark value

AMPS Hard

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #6 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 99%
Percentile: 98.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: AMPS_Hard. Category: Mathematics.

98.5% percentile inside its fair comparison set

99%Raw benchmark value

Integrals with game

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #46 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 69%
Percentile: 30.8%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: integrals_with_game. Category: Mathematics.

30.8% percentile inside its fair comparison set

69%Raw benchmark value

Math competition

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #31 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 96.1%
Percentile: 66.2%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: math_comp. Category: Mathematics.

66.2% percentile inside its fair comparison set

96.1%Raw benchmark value

Olympiad

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #52 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 87.5%
Percentile: 21.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: olympiad. Category: Mathematics.

21.5% percentile inside its fair comparison set

87.5%Raw benchmark value

Theory of mind

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #2 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 88.5%
Percentile: 100%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: theory_of_mind. Category: Reasoning.

100% percentile inside its fair comparison set

88.5%Raw benchmark value

Zebra puzzle

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #43 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 88.8%
Percentile: 35.4%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: zebra_puzzle. Category: Reasoning.

35.4% percentile inside its fair comparison set

88.8%Raw benchmark value

Spatial

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #51 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 96%
Percentile: 41.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: spatial. Category: Reasoning.

41.5% percentile inside its fair comparison set

96%Raw benchmark value

Logic with navigation

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #44 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 70%
Percentile: 44.6%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: logic_with_navigation. Category: Reasoning.

44.6% percentile inside its fair comparison set

70%Raw benchmark value

Professional reasoning33 benchmarks53.2%

Text Arena · Expert

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena expert leaderboard.

Rank #19 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,518
Percentile: 94.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: expert. Source rank: #19. Votes: 1797. Organization: zai. License: MIT.

94.5% percentile inside its fair comparison set

1,518Raw benchmark valueCI 1,504 - 1,532

Text Arena · Industry Business And Management And Financial Operations

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_business_and_management_and_financial_operations leaderboard.

Rank #30 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,475
Percentile: 92.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_business_and_management_and_financial_operations. Source rank: #31. Votes: 3346. Organization: zai. License: MIT.

92.1% percentile inside its fair comparison set

1,475Raw benchmark valueCI 1,464 - 1,486

Text Arena · Industry Entertainment And Sports And Media

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_entertainment_and_sports_and_media leaderboard.

Rank #36 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,447
Percentile: 90.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_entertainment_and_sports_and_media. Source rank: #38. Votes: 4756. Organization: zai. License: MIT.

90.6% percentile inside its fair comparison set

1,447Raw benchmark valueCI 1,437 - 1,456

Text Arena · Industry Legal And Government

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_legal_and_government leaderboard.

Rank #50 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,470
Percentile: 86%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_legal_and_government. Source rank: #55. Votes: 1578. Organization: zai. License: MIT.

86% percentile inside its fair comparison set

1,470Raw benchmark valueCI 1,455 - 1,486

Text Arena · Industry Life And Physical And Social Science

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_life_and_physical_and_social_science leaderboard.

Rank #19 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,505
Percentile: 95.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_life_and_physical_and_social_science. Source rank: #19. Votes: 2984. Organization: zai. License: MIT.

95.2% percentile inside its fair comparison set

1,505Raw benchmark valueCI 1,494 - 1,516

Text Arena · Industry Mathematical

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_mathematical leaderboard.

Rank #18 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,499
Percentile: 95.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_mathematical. Source rank: #18. Votes: 939. Organization: zai. License: MIT.

95.2% percentile inside its fair comparison set

1,499Raw benchmark valueCI 1,480 - 1,519

Text Arena · Industry Medicine And Healthcare

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_medicine_and_healthcare leaderboard.

Rank #25 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,497
Percentile: 93.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_medicine_and_healthcare. Source rank: #25. Votes: 1282. Organization: zai. License: MIT.

93.1% percentile inside its fair comparison set

1,497Raw benchmark valueCI 1,479 - 1,514

Text Arena · Industry Software And It Services

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_software_and_it_services leaderboard.

Rank #29 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,512
Percentile: 92.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_software_and_it_services. Source rank: #29. Votes: 6662. Organization: zai. License: MIT.

92.6% percentile inside its fair comparison set

1,512Raw benchmark valueCI 1,504 - 1,520

Text Arena · Industry Writing And Literature And Language

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_writing_and_literature_and_language leaderboard.

Rank #27 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,466
Percentile: 93.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_writing_and_literature_and_language. Source rank: #29. Votes: 5117. Organization: zai. License: MIT.

93.1% percentile inside its fair comparison set

1,466Raw benchmark valueCI 1,457 - 1,475

Text Arena · Expert · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena expert leaderboard.

Rank #17 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,514
Percentile: 95.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: expert. Source rank: #17. Votes: 1797. Organization: zai. License: MIT.

95.1% percentile inside its fair comparison set

1,514Raw benchmark valueCI 1,500 - 1,528

Text Arena · Industry Business And Management And Financial Operations · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_business_and_management_and_financial_operations leaderboard.

Rank #30 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,461
Percentile: 92.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_business_and_management_and_financial_operations. Source rank: #31. Votes: 3346. Organization: zai. License: MIT.

92.1% percentile inside its fair comparison set

1,461Raw benchmark valueCI 1,450 - 1,471

Text Arena · Industry Entertainment And Sports And Media · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_entertainment_and_sports_and_media leaderboard.

Rank #30 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,446
Percentile: 92.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_entertainment_and_sports_and_media. Source rank: #32. Votes: 4756. Organization: zai. License: MIT.

92.2% percentile inside its fair comparison set

1,446Raw benchmark valueCI 1,436 - 1,455

Text Arena · Industry Legal And Government · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_legal_and_government leaderboard.

Rank #39 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,469
Percentile: 89.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_legal_and_government. Source rank: #41. Votes: 1578. Organization: zai. License: MIT.

89.1% percentile inside its fair comparison set

1,469Raw benchmark valueCI 1,454 - 1,484

Text Arena · Industry Life And Physical And Social Science · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_life_and_physical_and_social_science leaderboard.

Rank #20 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,496
Percentile: 94.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_life_and_physical_and_social_science. Source rank: #20. Votes: 2984. Organization: zai. License: MIT.

94.9% percentile inside its fair comparison set

1,496Raw benchmark valueCI 1,485 - 1,507

Text Arena · Industry Mathematical · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_mathematical leaderboard.

Rank #13 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,499
Percentile: 96.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_mathematical. Source rank: #13. Votes: 939. Organization: zai. License: MIT.

96.6% percentile inside its fair comparison set

1,499Raw benchmark valueCI 1,480 - 1,518

Text Arena · Industry Medicine And Healthcare · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_medicine_and_healthcare leaderboard.

Rank #25 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,477
Percentile: 93.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_medicine_and_healthcare. Source rank: #26. Votes: 1282. Organization: zai. License: MIT.

93.1% percentile inside its fair comparison set

1,477Raw benchmark valueCI 1,460 - 1,494

Text Arena · Industry Software And It Services · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_software_and_it_services leaderboard.

Rank #29 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,492
Percentile: 92.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_software_and_it_services. Source rank: #29. Votes: 6662. Organization: zai. License: MIT.

92.6% percentile inside its fair comparison set

1,492Raw benchmark valueCI 1,484 - 1,500

Text Arena · Industry Writing And Literature And Language · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_writing_and_literature_and_language leaderboard.

Rank #21 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,467
Percentile: 94.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_writing_and_literature_and_language. Source rank: #22. Votes: 5117. Organization: zai. License: MIT.

94.7% percentile inside its fair comparison set

1,467Raw benchmark valueCI 1,458 - 1,476

Vals Index

VALS-AI · Professional reasoning · Combined

Weighted model performance across economically relevant Vals tasks.

Rank #16 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 54
Percentile: 62.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: vals_index; provider: Fireworks AI.

62.5% percentile inside its fair comparison set

54Raw benchmark valueCI 51 - 56

Legal Research Bench

VALS-AI · Professional reasoning · Objective

Applied legal research tasks.

Rank #7 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 49%
Percentile: 91.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: legal_research; provider: Zhipu AI.

91.2% percentile inside its fair comparison set

49%Raw benchmark valueCI 42.2% - 55.8%

Harvey's Legal Agent Benchmark

VALS-AI · Professional reasoning · Objective

Completing legal work with documents, spreadsheets, presentations, and file-system tools.

Rank #20 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 8.3%
Percentile: 73.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: hlab; provider: Zhipu AI.

73.9% percentile inside its fair comparison set

8.3%Raw benchmark valueCI 4.4% - 12.2%

LegalBench

VALS-AI · Professional reasoning · Objective

Academic legal reasoning tasks.

Rank #28 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 84.8%
Percentile: 80%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: legal_bench; provider: Zhipu AI.

80% percentile inside its fair comparison set

84.8%Raw benchmark valueCI 84.1% - 85.6%

Finance Agent v2

VALS-AI · Professional reasoning · Objective

Core financial analyst tasks for agentic models.

Rank #17 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 55.8%
Percentile: 76.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: fabv2; provider: Fireworks AI.

76.8% percentile inside its fair comparison set

55.8%Raw benchmark valueCI 51.8% - 59.9%

MedCode

VALS-AI · Professional reasoning · Objective

Medical billing support and coding tasks.

Rank #47 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 42.9%
Percentile: 51.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: medcode; provider: Zhipu AI.

51.6% percentile inside its fair comparison set

42.9%Raw benchmark valueCI 38.7% - 47%

MedScribe

VALS-AI · Professional reasoning · Objective

Administrative documentation support for doctors.

Rank #9 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 88.8%
Percentile: 91.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: medscribe; provider: Zhipu AI.

91.7% percentile inside its fair comparison set

88.8%Raw benchmark valueCI 84.9% - 92.7%

Public Benefits Bench

VALS-AI · Professional reasoning · Objective

Answering SNAP benefits questions across the public-benefits lifecycle.

Rank #8 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 68.5%
Percentile: 84.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: public-benefits-bench; provider: Zhipu AI.

84.4% percentile inside its fair comparison set

68.5%Raw benchmark valueCI 66.2% - 70.9%

SkillsBench

VALS-AI · Professional reasoning · Objective

Applied professional skills tasks.

Rank #24 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 47.5%
Percentile: 30.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: skillsbench; provider: Zhipu AI.

30.3% percentile inside its fair comparison set

47.5%Raw benchmark valueCI 38.6% - 56.4%

TaxEval v2

VALS-AI · Professional reasoning · Objective

Answer quality on tax questions and responses.

Rank #63 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 72.4%
Percentile: 51.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: tax_eval_v2; provider: Zhipu AI.

51.9% percentile inside its fair comparison set

72.4%Raw benchmark valueCI 70.6% - 74.1%

Data analysis

LB · Professional reasoning · Objective

Structured data manipulation and table reasoning accuracy.

Rank #51 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 70.2%
Percentile: 23.1%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Data Analysis. Tasks scored: 3.

23.1% percentile inside its fair comparison set

70.2%Raw benchmark value

Overall

LB · Professional reasoning · Objective

Average objective performance across LiveBench's current public category mix.

Rank #32 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 76.1%
Percentile: 52.3%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category averages included: 7.

52.3% percentile inside its fair comparison set

76.1%Raw benchmark value

Consecutive events

LB · Professional reasoning · Objective

Objective consecutive events score in LiveBench.

Rank #44 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 72.3%
Percentile: 33.8%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: consecutive_events. Category: Data Analysis.

33.8% percentile inside its fair comparison set

72.3%Raw benchmark value

Table join

LB · Professional reasoning · Objective

Objective table join score in LiveBench.

Rank #63 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 42.3%
Percentile: 4.6%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: tablejoin. Category: Data Analysis.

4.6% percentile inside its fair comparison set

42.3%Raw benchmark value

Table reformat

LB · Professional reasoning · Objective

Objective table reformat score in LiveBench.

Rank #58 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 96.1%
Percentile: 20%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: tablereformat. Category: Data Analysis.

20% percentile inside its fair comparison set

96.1%Raw benchmark value

Long context1 benchmark87%

Long Context Reasoning

AA · Long context · Objective

It checks whether long-context claims survive contact with retrieval, memory, or long-document tasks.

Rank #64 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 79.7%
Percentile: 87%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `lcr`.

87% percentile inside its fair comparison set

79.7%Raw benchmark value

Multilingual18 benchmarksn/a

Text Arena · Chinese

AR · Multilingual · Human

Observed user preference in Arena's Text Arena chinese leaderboard.

Rank #19 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,531
Percentile: 94.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: chinese. Source rank: #19. Votes: 1328. Organization: zai. License: MIT.

94.9% percentile inside its fair comparison set

1,531Raw benchmark valueCI 1,514 - 1,547

Text Arena · French

AR · Multilingual · Human

Observed user preference in Arena's Text Arena french leaderboard.

Rank #11 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,510
Percentile: 96.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: french. Source rank: #11. Votes: 532. Organization: zai. License: MIT.

96.1% percentile inside its fair comparison set

1,510Raw benchmark valueCI 1,483 - 1,536

Text Arena · German

AR · Multilingual · Human

Observed user preference in Arena's Text Arena german leaderboard.

Rank #2 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,520
Percentile: 99.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: german. Source rank: #2. Votes: 282. Organization: zai. License: MIT.

99.6% percentile inside its fair comparison set

1,520Raw benchmark valueCI 1,486 - 1,554

Text Arena · Japanese

AR · Multilingual · Human

Observed user preference in Arena's Text Arena japanese leaderboard.

Rank #32 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,454
Percentile: 87.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: japanese. Source rank: #34. Votes: 327. Organization: zai. License: MIT.

87.2% percentile inside its fair comparison set

1,454Raw benchmark valueCI 1,420 - 1,488

Text Arena · Korean

AR · Multilingual · Human

Observed user preference in Arena's Text Arena korean leaderboard.

Rank #6 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,481
Percentile: 98%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: korean. Source rank: #6. Votes: 298. Organization: zai. License: MIT.

98% percentile inside its fair comparison set

1,481Raw benchmark valueCI 1,446 - 1,515

Text Arena · Non English

AR · Multilingual · Human

Observed user preference in Arena's Text Arena non_english leaderboard.

Rank #34 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,463
Percentile: 91.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: non_english. Source rank: #36. Votes: 10629. Organization: zai. License: MIT.

91.2% percentile inside its fair comparison set

1,463Raw benchmark valueCI 1,456 - 1,470

Text Arena · Polish

AR · Multilingual · Human

Observed user preference in Arena's Text Arena polish leaderboard.

Rank #16 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,492
Percentile: 92.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: polish. Source rank: #17. Votes: 248. Organization: zai. License: MIT.

92.5% percentile inside its fair comparison set

1,492Raw benchmark valueCI 1,456 - 1,529

Text Arena · Russian

AR · Multilingual · Human

Observed user preference in Arena's Text Arena russian leaderboard.

Rank #45 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,467
Percentile: 87.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: russian. Source rank: #51. Votes: 2211. Organization: zai. License: MIT.

87.1% percentile inside its fair comparison set

1,467Raw benchmark valueCI 1,454 - 1,480

Text Arena · Spanish

AR · Multilingual · Human

Observed user preference in Arena's Text Arena spanish leaderboard.

Rank #64 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,448
Percentile: 75.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: spanish. Source rank: #70. Votes: 538. Organization: zai. License: MIT.

75.7% percentile inside its fair comparison set

1,448Raw benchmark valueCI 1,424 - 1,473

Text Arena · Chinese · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena chinese leaderboard.

Rank #25 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,524
Percentile: 93.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: chinese. Source rank: #27. Votes: 1328. Organization: zai. License: MIT.

93.2% percentile inside its fair comparison set

1,524Raw benchmark valueCI 1,507 - 1,541

Text Arena · French · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena french leaderboard.

Rank #17 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,494
Percentile: 93.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: french. Source rank: #17. Votes: 532. Organization: zai. License: MIT.

93.8% percentile inside its fair comparison set

1,494Raw benchmark valueCI 1,468 - 1,520

Text Arena · German · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena german leaderboard.

Rank #5 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,504
Percentile: 98.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: german. Source rank: #5. Votes: 282. Organization: zai. License: MIT.

98.5% percentile inside its fair comparison set

1,504Raw benchmark valueCI 1,470 - 1,538

Text Arena · Japanese · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena japanese leaderboard.

Rank #30 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,445
Percentile: 88.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: japanese. Source rank: #33. Votes: 327. Organization: zai. License: MIT.

88.1% percentile inside its fair comparison set

1,445Raw benchmark valueCI 1,411 - 1,480

Text Arena · Korean · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena korean leaderboard.

Rank #7 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,475
Percentile: 97.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: korean. Source rank: #7. Votes: 298. Organization: zai. License: MIT.

97.5% percentile inside its fair comparison set

1,475Raw benchmark valueCI 1,441 - 1,510

Text Arena · Non English · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena non_english leaderboard.

Rank #31 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,457
Percentile: 92%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: non_english. Source rank: #33. Votes: 10629. Organization: zai. License: MIT.

92% percentile inside its fair comparison set

1,457Raw benchmark valueCI 1,450 - 1,463

Text Arena · Polish · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena polish leaderboard.

Rank #16 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,484
Percentile: 92.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: polish. Source rank: #17. Votes: 248. Organization: zai. License: MIT.

92.5% percentile inside its fair comparison set

1,484Raw benchmark valueCI 1,448 - 1,521

Text Arena · Russian · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena russian leaderboard.

Rank #38 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,458
Percentile: 89.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: russian. Source rank: #40. Votes: 2211. Organization: zai. License: MIT.

89.1% percentile inside its fair comparison set

1,458Raw benchmark valueCI 1,445 - 1,471

Text Arena · Spanish · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena spanish leaderboard.

Rank #63 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,446
Percentile: 76.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: spanish. Source rank: #66. Votes: 538. Organization: zai. License: MIT.

76.1% percentile inside its fair comparison set

1,446Raw benchmark valueCI 1,421 - 1,470

Source links and registry checks

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Artificial Analysis

Oct 3, 2026

source →

official

LiveBench

Oct 3, 2026

source →

official

Terminal-Bench

Oct 3, 2026

source →

Model profile · Zhipu

glm-5.3-max

Closed weightsmid · registry tag 2026 benchmark-derived

Thin verified coverage

Reads as thin verified coverage across the resolved source data. Capability fit excludes speed and cost.

Visible coverage: 45.9%
Verified coverage: 45.9%
Spread: 95.4%
Last verified: Oct 3, 2026

72%capability fit

textcodedocument5 aliases35 official source links

Open compare

Data version

Current snapshot.

Published Oct 3, 2026Model list checked4 stale sources105 claim warnings9 providers · 1334 tracked models

The registry snapshot and page stamp are shown so a stale deploy is visible at a glance. View live data status.

Source-linked scores by benchmark

Thin verified coverageThis model currently reads as thin verified coverage across the resolved source data.

Chat / text35 benchmarks81.4%

Intelligence Index

AA · Chat / text · Combined

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #34 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 45
Percentile: 94.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `intelligenceIndex`.

94.1% percentile inside its fair comparison set

45Raw benchmark value

AA-Omniscience accuracy

AA · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #109 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 33.9%
Percentile: 76.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `omniscienceAccuracy`.

76.2% percentile inside its fair comparison set

33.9%Raw benchmark value

AA-Omniscience non-hallucination

AA · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #29 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 70.4%
Percentile: 93.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `omniscienceNonHallucination`.

93.8% percentile inside its fair comparison set

70.4%Raw benchmark value

Input price

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #252 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $1.4 /1M input tokens
Percentile: 29.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mInputTokens`.

29.4% percentile inside its fair comparison set

$1.4 /1M input tokensRaw benchmark value

Output price

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #241 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: $4.4 /1M output tokens
Percentile: 33.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `price1mOutputTokens`.

33.3% percentile inside its fair comparison set

$4.4 /1M output tokensRaw benchmark value

Output Speed

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #162 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 70.7 tokens/s
Percentile: 38.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianOutputTokensPerSecond`.

38.5% percentile inside its fair comparison set

70.7 tokens/sRaw benchmark value

Time to first token

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #153 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 3.42s
Percentile: 42.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianTimeToFirstTokenSeconds`.

42.4% percentile inside its fair comparison set

3.42sRaw benchmark value

Time to first answer token

AA · Chat / text · Speed / cost

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #180 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 31.70s
Percentile: 31.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `medianTimeToFirstAnswerTokenSeconds`.

31.7% percentile inside its fair comparison set

31.70sRaw benchmark value

Text Arena

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #27 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,478
Percentile: 93.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: overall. Source rank: #27. Votes: 17857. Organization: zai. License: MIT.

93.1% percentile inside its fair comparison set

1,478Raw benchmark valueCI 1,473 - 1,484

Text Arena · Creative Writing

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #33 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,454
Percentile: 91.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: creative_writing. Source rank: #34. Votes: 4034. Organization: zai. License: MIT.

91.4% percentile inside its fair comparison set

1,454Raw benchmark valueCI 1,444 - 1,464

Text Arena · English

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #23 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,490
Percentile: 94.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: english. Source rank: #23. Votes: 7223. Organization: zai. License: MIT.

94.1% percentile inside its fair comparison set

1,490Raw benchmark valueCI 1,482 - 1,497

Text Arena · Exclude Ties

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #28 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,488
Percentile: 92.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: exclude_ties. Source rank: #28. Votes: 13198. Organization: zai. License: MIT.

92.8% percentile inside its fair comparison set

1,488Raw benchmark valueCI 1,481 - 1,495

Text Arena · Hard Prompts

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #25 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,503
Percentile: 93.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: hard_prompts. Source rank: #25. Votes: 11434. Organization: zai. License: MIT.

93.6% percentile inside its fair comparison set

1,503Raw benchmark valueCI 1,497 - 1,510

Text Arena · Hard Prompts English

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #18 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,512
Percentile: 95.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: hard_prompts_english. Source rank: #18. Votes: 4618. Organization: zai. License: MIT.

95.5% percentile inside its fair comparison set

1,512Raw benchmark valueCI 1,503 - 1,521

Text Arena · Instruction Following

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #25 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,476
Percentile: 93.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: instruction_following. Source rank: #25. Votes: 6049. Organization: zai. License: MIT.

93.6% percentile inside its fair comparison set

1,476Raw benchmark valueCI 1,467 - 1,484

Text Arena · Longer Query

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #33 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,486
Percentile: 91%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: longer_query. Source rank: #34. Votes: 8337. Organization: zai. License: MIT.

91% percentile inside its fair comparison set

1,486Raw benchmark valueCI 1,479 - 1,494

Text Arena · Multi Turn

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #39 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,480
Percentile: 89.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: multi_turn. Source rank: #41. Votes: 2694. Organization: zai. License: MIT.

89.8% percentile inside its fair comparison set

1,480Raw benchmark valueCI 1,468 - 1,492

Text Arena · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #25 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,471
Percentile: 93.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: overall. Source rank: #26. Votes: 17857. Organization: zai. License: MIT.

93.6% percentile inside its fair comparison set

1,471Raw benchmark valueCI 1,466 - 1,477

Text Arena · Creative Writing · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #23 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,457
Percentile: 94.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: creative_writing. Source rank: #24. Votes: 4034. Organization: zai. License: MIT.

94.1% percentile inside its fair comparison set

1,457Raw benchmark valueCI 1,447 - 1,467

Text Arena · English · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #19 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,482
Percentile: 95.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: english. Source rank: #19. Votes: 7223. Organization: zai. License: MIT.

95.2% percentile inside its fair comparison set

1,482Raw benchmark valueCI 1,474 - 1,490

Text Arena · Exclude Ties · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #26 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,477
Percentile: 93.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: exclude_ties. Source rank: #27. Votes: 13198. Organization: zai. License: MIT.

93.4% percentile inside its fair comparison set

1,477Raw benchmark valueCI 1,469 - 1,484

Text Arena · Hard Prompts · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #21 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,488
Percentile: 94.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: hard_prompts. Source rank: #21. Votes: 11434. Organization: zai. License: MIT.

94.7% percentile inside its fair comparison set

1,488Raw benchmark valueCI 1,482 - 1,495

Text Arena · Hard Prompts English · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #16 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,497
Percentile: 96%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: hard_prompts_english. Source rank: #16. Votes: 4618. Organization: zai. License: MIT.

96% percentile inside its fair comparison set

1,497Raw benchmark valueCI 1,488 - 1,506

Text Arena · Instruction Following · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #23 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,477
Percentile: 94.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: instruction_following. Source rank: #23. Votes: 6049. Organization: zai. License: MIT.

94.1% percentile inside its fair comparison set

1,477Raw benchmark valueCI 1,469 - 1,485

Text Arena · Longer Query · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #20 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,483
Percentile: 94.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: longer_query. Source rank: #20. Votes: 8337. Organization: zai. License: MIT.

94.6% percentile inside its fair comparison set

1,483Raw benchmark valueCI 1,476 - 1,490

Text Arena · Multi Turn · No Style Control

AR · Chat / text · Human

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #31 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,472
Percentile: 92%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: multi_turn. Source rank: #33. Votes: 2694. Organization: zai. License: MIT.

92% percentile inside its fair comparison set

1,472Raw benchmark valueCI 1,460 - 1,484

Instruction following

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #33 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 69.3%
Percentile: 50.8%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: IF. Tasks scored: 4.

50.8% percentile inside its fair comparison set

69.3%Raw benchmark value

Language

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #31 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 79.9%
Percentile: 53.8%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Language. Tasks scored: 3.

53.8% percentile inside its fair comparison set

79.9%Raw benchmark value

Paraphrase

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #40 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 64.7%
Percentile: 40%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: paraphrase. Category: IF.

40% percentile inside its fair comparison set

64.7%Raw benchmark value

Simplify

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #29 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 65.9%
Percentile: 56.9%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: simplify. Category: IF.

56.9% percentile inside its fair comparison set

65.9%Raw benchmark value

Story generation

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #31 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 72.1%
Percentile: 53.8%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: story_generation. Category: IF.

53.8% percentile inside its fair comparison set

72.1%Raw benchmark value

Summarize

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #22 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 74.6%
Percentile: 67.7%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: summarize. Category: IF.

67.7% percentile inside its fair comparison set

74.6%Raw benchmark value

Connections

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #45 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 95.3%
Percentile: 32.3%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: connections. Category: Language.

32.3% percentile inside its fair comparison set

95.3%Raw benchmark value

Plot unscrambling

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #37 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 60.2%
Percentile: 44.6%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: plot_unscrambling. Category: Language.

44.6% percentile inside its fair comparison set

60.2%Raw benchmark value

Typos

LB · Chat / text · Objective

It tests whether the model is actually useful in normal conversational turns, not just on narrow correctness tasks.

Rank #23 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 84%
Percentile: 80%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: typos. Category: Language.

80% percentile inside its fair comparison set

84%Raw benchmark value

Coding21 benchmarks73.6%

SciCode

AA · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #14 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 59%
Percentile: 93.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `scicode`.

93.4% percentile inside its fair comparison set

59%Raw benchmark value

Text Arena · Coding

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #34 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,521
Percentile: 91.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: coding. Source rank: #34. Votes: 4424. Organization: zai. License: MIT.

91.1% percentile inside its fair comparison set

1,521Raw benchmark valueCI 1,511 - 1,530

Text Arena · Coding · No Style Control

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #27 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,497
Percentile: 93%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: coding. Source rank: #27. Votes: 4424. Organization: zai. License: MIT.

93% percentile inside its fair comparison set

1,497Raw benchmark valueCI 1,487 - 1,506

Vibe Code Bench v1.1

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #23 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 78.1%
Percentile: 76.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: vibe-code; provider: Zhipu AI.

76.3% percentile inside its fair comparison set

78.1%Raw benchmark valueCI 70.4% - 85.8%

WebDev Arena

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #19 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,623
Percentile: 85.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: overall. Source rank: #20. Votes: 7658. Organization: zai. License: MIT.

85.8% percentile inside its fair comparison set

1,623Raw benchmark valueCI 1,615 - 1,631

Code Arena · Webdev Html

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #25 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,590
Percentile: 80.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: webdev-html. Source rank: #26. Votes: 1016. Organization: zai. License: MIT.

80.8% percentile inside its fair comparison set

1,590Raw benchmark valueCI 1,569 - 1,610

Code Arena · Webdev React

AR · Coding · Human

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #17 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,641
Percentile: 85.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: webdev-react. Source rank: #18. Votes: 5094. Organization: zai. License: MIT.

85.8% percentile inside its fair comparison set

1,641Raw benchmark valueCI 1,631 - 1,651

Code Migration

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #21 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 44.2%
Percentile: 70.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: code-migration; provider: Zhipu AI.

70.6% percentile inside its fair comparison set

44.2%Raw benchmark valueCI 35.8% - 52.6%

IOI

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #12 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 68.4%
Percentile: 70.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: ioi; provider: Zhipu AI.

70.3% percentile inside its fair comparison set

68.4%Raw benchmark valueCI 53.8% - 83.1%

ProgramBench

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #13 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 1.5%
Percentile: 81%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: programbench; provider: Zhipu AI.

81% percentile inside its fair comparison set

1.5%Raw benchmark valueCI 0% - 3.2%

Terminal-Bench 2.1

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #25 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 71.5%
Percentile: 64.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: terminal-bench-2-1; provider: Zhipu AI.

64.2% percentile inside its fair comparison set

71.5%Raw benchmark valueCI 67.7% - 75.4%

SWE-bench Verified

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #6 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 95.4%
Percentile: 95.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: swebench; provider: Zhipu AI.

95.1% percentile inside its fair comparison set

95.4%Raw benchmark valueCI 93.6% - 97.2%

LiveCodeBench

VALS-AI · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #65 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 80.5%
Percentile: 48.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: lcb; provider: Zhipu AI.

48.8% percentile inside its fair comparison set

80.5%Raw benchmark valueCI 78.4% - 82.6%

Terminal-Bench 4.0

TERMINAL-BENCH · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #16 · Source label: GLM-5.3 (max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Terminal-Bench
Raw value: 41.8%
Percentile: 42.3%
Last updated: aging
Eligibility: headline eligible

42.3% percentile inside its fair comparison set

41.8%Raw benchmark valueCI 38.6% - 45.1%

Agentic coding

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #15 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 60.9%
Percentile: 78.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Agentic Coding. Tasks scored: 3.

78.5% percentile inside its fair comparison set

60.9%Raw benchmark value

Coding

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #26 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 79%
Percentile: 61.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Coding. Tasks scored: 2.

61.5% percentile inside its fair comparison set

79%Raw benchmark value

JavaScript

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #12 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 72.7%
Percentile: 87.7%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: javascript. Category: Agentic Coding.

87.7% percentile inside its fair comparison set

72.7%Raw benchmark value

TypeScript

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #17 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 50%
Percentile: 81.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: typescript. Category: Agentic Coding.

81.5% percentile inside its fair comparison set

50%Raw benchmark value

Python

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #21 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 60%
Percentile: 80%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: python. Category: Agentic Coding.

80% percentile inside its fair comparison set

60%Raw benchmark value

Coding generation

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #35 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 77.5%
Percentile: 52.3%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: code_generation. Category: Coding.

52.3% percentile inside its fair comparison set

77.5%Raw benchmark value

Coding completion

LB · Coding · Objective

It tells you whether the model can generate, repair, and reason over code under evaluator pressure rather than marketing examples.

Rank #23 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 80.4%
Percentile: 78.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: code_completion. Category: Coding.

78.5% percentile inside its fair comparison set

80.4%Raw benchmark value

Reasoning / math / science18 benchmarks68.1%

Humanity's Last Exam

AA · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #53 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 42.3%
Percentile: 90.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `hle`.

90.2% percentile inside its fair comparison set

42.3%Raw benchmark value

GPQA

AA · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #35 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 91.7%
Percentile: 93.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `gpqa`.

93.1% percentile inside its fair comparison set

91.7%Raw benchmark value

CritPt

AA · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #47 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 19.1%
Percentile: 89.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `critpt`.

89.9% percentile inside its fair comparison set

19.1%Raw benchmark value

Text Arena · Math

AR · Reasoning / math / science · Human

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #20 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,494
Percentile: 94.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: math. Source rank: #20. Votes: 739. Organization: zai. License: MIT.

94.7% percentile inside its fair comparison set

1,494Raw benchmark valueCI 1,473 - 1,515

Text Arena · Math · No Style Control

AR · Reasoning / math / science · Human

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #19 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,491
Percentile: 95%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: math. Source rank: #19. Votes: 739. Organization: zai. License: MIT.

95% percentile inside its fair comparison set

1,491Raw benchmark valueCI 1,470 - 1,512

ProofBench

VALS-AI · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #26 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 49%
Percentile: 40.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: proof_bench; provider: Zhipu AI.

40.5% percentile inside its fair comparison set

49%Raw benchmark valueCI 39.2% - 58.8%

GPQA Diamond

VALS-AI · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #36 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 88.1%
Percentile: 72.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: gpqa; provider: Zhipu AI.

72.6% percentile inside its fair comparison set

88.1%Raw benchmark valueCI 84.5% - 91.8%

MMLU Pro

VALS-AI · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #38 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 86.8%
Percentile: 70.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: mmlu_pro; provider: Zhipu AI.

70.4% percentile inside its fair comparison set

86.8%Raw benchmark valueCI 86.1% - 87.4%

Mathematics

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #43 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 87.9%
Percentile: 35.4%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Mathematics. Tasks scored: 4.

35.4% percentile inside its fair comparison set

87.9%Raw benchmark value

Reasoning

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #35 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 85.8%
Percentile: 47.7%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Reasoning. Tasks scored: 4.

47.7% percentile inside its fair comparison set

85.8%Raw benchmark value

AMPS Hard

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #6 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 99%
Percentile: 98.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: AMPS_Hard. Category: Mathematics.

98.5% percentile inside its fair comparison set

99%Raw benchmark value

Integrals with game

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #46 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 69%
Percentile: 30.8%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: integrals_with_game. Category: Mathematics.

30.8% percentile inside its fair comparison set

69%Raw benchmark value

Math competition

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #31 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 96.1%
Percentile: 66.2%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: math_comp. Category: Mathematics.

66.2% percentile inside its fair comparison set

96.1%Raw benchmark value

Olympiad

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #52 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 87.5%
Percentile: 21.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: olympiad. Category: Mathematics.

21.5% percentile inside its fair comparison set

87.5%Raw benchmark value

Theory of mind

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #2 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 88.5%
Percentile: 100%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: theory_of_mind. Category: Reasoning.

100% percentile inside its fair comparison set

88.5%Raw benchmark value

Zebra puzzle

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #43 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 88.8%
Percentile: 35.4%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: zebra_puzzle. Category: Reasoning.

35.4% percentile inside its fair comparison set

88.8%Raw benchmark value

Spatial

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #51 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 96%
Percentile: 41.5%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: spatial. Category: Reasoning.

41.5% percentile inside its fair comparison set

96%Raw benchmark value

Logic with navigation

LB · Reasoning / math / science · Objective

It is one of the cleaner reads on deliberate reasoning strength rather than style or popularity.

Rank #44 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 70%
Percentile: 44.6%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: logic_with_navigation. Category: Reasoning.

44.6% percentile inside its fair comparison set

70%Raw benchmark value

Professional reasoning33 benchmarks53.2%

Text Arena · Expert

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena expert leaderboard.

Rank #19 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,518
Percentile: 94.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: expert. Source rank: #19. Votes: 1797. Organization: zai. License: MIT.

94.5% percentile inside its fair comparison set

1,518Raw benchmark valueCI 1,504 - 1,532

Text Arena · Industry Business And Management And Financial Operations

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_business_and_management_and_financial_operations leaderboard.

Rank #30 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,475
Percentile: 92.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_business_and_management_and_financial_operations. Source rank: #31. Votes: 3346. Organization: zai. License: MIT.

92.1% percentile inside its fair comparison set

1,475Raw benchmark valueCI 1,464 - 1,486

Text Arena · Industry Entertainment And Sports And Media

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_entertainment_and_sports_and_media leaderboard.

Rank #36 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,447
Percentile: 90.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_entertainment_and_sports_and_media. Source rank: #38. Votes: 4756. Organization: zai. License: MIT.

90.6% percentile inside its fair comparison set

1,447Raw benchmark valueCI 1,437 - 1,456

Text Arena · Industry Legal And Government

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_legal_and_government leaderboard.

Rank #50 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,470
Percentile: 86%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_legal_and_government. Source rank: #55. Votes: 1578. Organization: zai. License: MIT.

86% percentile inside its fair comparison set

1,470Raw benchmark valueCI 1,455 - 1,486

Text Arena · Industry Life And Physical And Social Science

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_life_and_physical_and_social_science leaderboard.

Rank #19 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,505
Percentile: 95.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_life_and_physical_and_social_science. Source rank: #19. Votes: 2984. Organization: zai. License: MIT.

95.2% percentile inside its fair comparison set

1,505Raw benchmark valueCI 1,494 - 1,516

Text Arena · Industry Mathematical

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_mathematical leaderboard.

Rank #18 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,499
Percentile: 95.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_mathematical. Source rank: #18. Votes: 939. Organization: zai. License: MIT.

95.2% percentile inside its fair comparison set

1,499Raw benchmark valueCI 1,480 - 1,519

Text Arena · Industry Medicine And Healthcare

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_medicine_and_healthcare leaderboard.

Rank #25 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,497
Percentile: 93.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_medicine_and_healthcare. Source rank: #25. Votes: 1282. Organization: zai. License: MIT.

93.1% percentile inside its fair comparison set

1,497Raw benchmark valueCI 1,479 - 1,514

Text Arena · Industry Software And It Services

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_software_and_it_services leaderboard.

Rank #29 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,512
Percentile: 92.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_software_and_it_services. Source rank: #29. Votes: 6662. Organization: zai. License: MIT.

92.6% percentile inside its fair comparison set

1,512Raw benchmark valueCI 1,504 - 1,520

Text Arena · Industry Writing And Literature And Language

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_writing_and_literature_and_language leaderboard.

Rank #27 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,466
Percentile: 93.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_writing_and_literature_and_language. Source rank: #29. Votes: 5117. Organization: zai. License: MIT.

93.1% percentile inside its fair comparison set

1,466Raw benchmark valueCI 1,457 - 1,475

Text Arena · Expert · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena expert leaderboard.

Rank #17 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,514
Percentile: 95.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: expert. Source rank: #17. Votes: 1797. Organization: zai. License: MIT.

95.1% percentile inside its fair comparison set

1,514Raw benchmark valueCI 1,500 - 1,528

Text Arena · Industry Business And Management And Financial Operations · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_business_and_management_and_financial_operations leaderboard.

Rank #30 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,461
Percentile: 92.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_business_and_management_and_financial_operations. Source rank: #31. Votes: 3346. Organization: zai. License: MIT.

92.1% percentile inside its fair comparison set

1,461Raw benchmark valueCI 1,450 - 1,471

Text Arena · Industry Entertainment And Sports And Media · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_entertainment_and_sports_and_media leaderboard.

Rank #30 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,446
Percentile: 92.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_entertainment_and_sports_and_media. Source rank: #32. Votes: 4756. Organization: zai. License: MIT.

92.2% percentile inside its fair comparison set

1,446Raw benchmark valueCI 1,436 - 1,455

Text Arena · Industry Legal And Government · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_legal_and_government leaderboard.

Rank #39 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,469
Percentile: 89.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_legal_and_government. Source rank: #41. Votes: 1578. Organization: zai. License: MIT.

89.1% percentile inside its fair comparison set

1,469Raw benchmark valueCI 1,454 - 1,484

Text Arena · Industry Life And Physical And Social Science · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_life_and_physical_and_social_science leaderboard.

Rank #20 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,496
Percentile: 94.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_life_and_physical_and_social_science. Source rank: #20. Votes: 2984. Organization: zai. License: MIT.

94.9% percentile inside its fair comparison set

1,496Raw benchmark valueCI 1,485 - 1,507

Text Arena · Industry Mathematical · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_mathematical leaderboard.

Rank #13 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,499
Percentile: 96.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_mathematical. Source rank: #13. Votes: 939. Organization: zai. License: MIT.

96.6% percentile inside its fair comparison set

1,499Raw benchmark valueCI 1,480 - 1,518

Text Arena · Industry Medicine And Healthcare · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_medicine_and_healthcare leaderboard.

Rank #25 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,477
Percentile: 93.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_medicine_and_healthcare. Source rank: #26. Votes: 1282. Organization: zai. License: MIT.

93.1% percentile inside its fair comparison set

1,477Raw benchmark valueCI 1,460 - 1,494

Text Arena · Industry Software And It Services · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_software_and_it_services leaderboard.

Rank #29 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,492
Percentile: 92.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_software_and_it_services. Source rank: #29. Votes: 6662. Organization: zai. License: MIT.

92.6% percentile inside its fair comparison set

1,492Raw benchmark valueCI 1,484 - 1,500

Text Arena · Industry Writing And Literature And Language · No Style Control

AR · Professional reasoning · Human

Observed user preference in Arena's Text Arena industry_writing_and_literature_and_language leaderboard.

Rank #21 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,467
Percentile: 94.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: industry_writing_and_literature_and_language. Source rank: #22. Votes: 5117. Organization: zai. License: MIT.

94.7% percentile inside its fair comparison set

1,467Raw benchmark valueCI 1,458 - 1,476

Vals Index

VALS-AI · Professional reasoning · Combined

Weighted model performance across economically relevant Vals tasks.

Rank #16 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 54
Percentile: 62.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: vals_index; provider: Fireworks AI.

62.5% percentile inside its fair comparison set

54Raw benchmark valueCI 51 - 56

Legal Research Bench

VALS-AI · Professional reasoning · Objective

Applied legal research tasks.

Rank #7 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 49%
Percentile: 91.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: legal_research; provider: Zhipu AI.

91.2% percentile inside its fair comparison set

49%Raw benchmark valueCI 42.2% - 55.8%

Harvey's Legal Agent Benchmark

VALS-AI · Professional reasoning · Objective

Completing legal work with documents, spreadsheets, presentations, and file-system tools.

Rank #20 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 8.3%
Percentile: 73.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: hlab; provider: Zhipu AI.

73.9% percentile inside its fair comparison set

8.3%Raw benchmark valueCI 4.4% - 12.2%

LegalBench

VALS-AI · Professional reasoning · Objective

Academic legal reasoning tasks.

Rank #28 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 84.8%
Percentile: 80%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: legal_bench; provider: Zhipu AI.

80% percentile inside its fair comparison set

84.8%Raw benchmark valueCI 84.1% - 85.6%

Finance Agent v2

VALS-AI · Professional reasoning · Objective

Core financial analyst tasks for agentic models.

Rank #17 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 55.8%
Percentile: 76.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: fabv2; provider: Fireworks AI.

76.8% percentile inside its fair comparison set

55.8%Raw benchmark valueCI 51.8% - 59.9%

MedCode

VALS-AI · Professional reasoning · Objective

Medical billing support and coding tasks.

Rank #47 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 42.9%
Percentile: 51.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: medcode; provider: Zhipu AI.

51.6% percentile inside its fair comparison set

42.9%Raw benchmark valueCI 38.7% - 47%

MedScribe

VALS-AI · Professional reasoning · Objective

Administrative documentation support for doctors.

Rank #9 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 88.8%
Percentile: 91.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: medscribe; provider: Zhipu AI.

91.7% percentile inside its fair comparison set

88.8%Raw benchmark valueCI 84.9% - 92.7%

Public Benefits Bench

VALS-AI · Professional reasoning · Objective

Answering SNAP benefits questions across the public-benefits lifecycle.

Rank #8 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 68.5%
Percentile: 84.4%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: public-benefits-bench; provider: Zhipu AI.

84.4% percentile inside its fair comparison set

68.5%Raw benchmark valueCI 66.2% - 70.9%

SkillsBench

VALS-AI · Professional reasoning · Objective

Applied professional skills tasks.

Rank #24 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 47.5%
Percentile: 30.3%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: skillsbench; provider: Zhipu AI.

30.3% percentile inside its fair comparison set

47.5%Raw benchmark valueCI 38.6% - 56.4%

TaxEval v2

VALS-AI · Professional reasoning · Objective

Answer quality on tax questions and responses.

Rank #63 · Source label: zai/glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Vals AI
Raw value: 72.4%
Percentile: 51.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Vals AI BenchmarkView overall scores. Vals slug: tax_eval_v2; provider: Zhipu AI.

51.9% percentile inside its fair comparison set

72.4%Raw benchmark valueCI 70.6% - 74.1%

Data analysis

LB · Professional reasoning · Objective

Structured data manipulation and table reasoning accuracy.

Rank #51 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 70.2%
Percentile: 23.1%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category: Data Analysis. Tasks scored: 3.

23.1% percentile inside its fair comparison set

70.2%Raw benchmark value

Overall

LB · Professional reasoning · Objective

Average objective performance across LiveBench's current public category mix.

Rank #32 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 76.1%
Percentile: 52.3%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Category averages included: 7.

52.3% percentile inside its fair comparison set

76.1%Raw benchmark value

Consecutive events

LB · Professional reasoning · Objective

Objective consecutive events score in LiveBench.

Rank #44 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 72.3%
Percentile: 33.8%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: consecutive_events. Category: Data Analysis.

33.8% percentile inside its fair comparison set

72.3%Raw benchmark value

Table join

LB · Professional reasoning · Objective

Objective table join score in LiveBench.

Rank #63 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 42.3%
Percentile: 4.6%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: tablejoin. Category: Data Analysis.

4.6% percentile inside its fair comparison set

42.3%Raw benchmark value

Table reformat

LB · Professional reasoning · Objective

Objective table reformat score in LiveBench.

Rank #58 · Source label: glm-5.3

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: LiveBench
Raw value: 96.1%
Percentile: 20%
Last updated: stale
Eligibility: headline eligible

Derived from the official LiveBench website leaderboard table. Task: tablereformat. Category: Data Analysis.

20% percentile inside its fair comparison set

96.1%Raw benchmark value

Long context1 benchmark87%

Long Context Reasoning

AA · Long context · Objective

It checks whether long-context claims survive contact with retrieval, memory, or long-document tasks.

Rank #64 · Source label: GLM-5.3 (Max)

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Artificial Analysis
Raw value: 79.7%
Percentile: 87%
Last updated: recent
Eligibility: headline eligible

Parsed from Artificial Analysis public leaderboard field `lcr`.

87% percentile inside its fair comparison set

79.7%Raw benchmark value

Multilingual18 benchmarksn/a

Text Arena · Chinese

AR · Multilingual · Human

Observed user preference in Arena's Text Arena chinese leaderboard.

Rank #19 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,531
Percentile: 94.9%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: chinese. Source rank: #19. Votes: 1328. Organization: zai. License: MIT.

94.9% percentile inside its fair comparison set

1,531Raw benchmark valueCI 1,514 - 1,547

Text Arena · French

AR · Multilingual · Human

Observed user preference in Arena's Text Arena french leaderboard.

Rank #11 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,510
Percentile: 96.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: french. Source rank: #11. Votes: 532. Organization: zai. License: MIT.

96.1% percentile inside its fair comparison set

1,510Raw benchmark valueCI 1,483 - 1,536

Text Arena · German

AR · Multilingual · Human

Observed user preference in Arena's Text Arena german leaderboard.

Rank #2 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,520
Percentile: 99.6%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: german. Source rank: #2. Votes: 282. Organization: zai. License: MIT.

99.6% percentile inside its fair comparison set

1,520Raw benchmark valueCI 1,486 - 1,554

Text Arena · Japanese

AR · Multilingual · Human

Observed user preference in Arena's Text Arena japanese leaderboard.

Rank #32 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,454
Percentile: 87.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: japanese. Source rank: #34. Votes: 327. Organization: zai. License: MIT.

87.2% percentile inside its fair comparison set

1,454Raw benchmark valueCI 1,420 - 1,488

Text Arena · Korean

AR · Multilingual · Human

Observed user preference in Arena's Text Arena korean leaderboard.

Rank #6 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,481
Percentile: 98%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: korean. Source rank: #6. Votes: 298. Organization: zai. License: MIT.

98% percentile inside its fair comparison set

1,481Raw benchmark valueCI 1,446 - 1,515

Text Arena · Non English

AR · Multilingual · Human

Observed user preference in Arena's Text Arena non_english leaderboard.

Rank #34 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,463
Percentile: 91.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: non_english. Source rank: #36. Votes: 10629. Organization: zai. License: MIT.

91.2% percentile inside its fair comparison set

1,463Raw benchmark valueCI 1,456 - 1,470

Text Arena · Polish

AR · Multilingual · Human

Observed user preference in Arena's Text Arena polish leaderboard.

Rank #16 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,492
Percentile: 92.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: polish. Source rank: #17. Votes: 248. Organization: zai. License: MIT.

92.5% percentile inside its fair comparison set

1,492Raw benchmark valueCI 1,456 - 1,529

Text Arena · Russian

AR · Multilingual · Human

Observed user preference in Arena's Text Arena russian leaderboard.

Rank #45 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,467
Percentile: 87.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: russian. Source rank: #51. Votes: 2211. Organization: zai. License: MIT.

87.1% percentile inside its fair comparison set

1,467Raw benchmark valueCI 1,454 - 1,480

Text Arena · Spanish

AR · Multilingual · Human

Observed user preference in Arena's Text Arena spanish leaderboard.

Rank #64 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,448
Percentile: 75.7%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: spanish. Source rank: #70. Votes: 538. Organization: zai. License: MIT.

75.7% percentile inside its fair comparison set

1,448Raw benchmark valueCI 1,424 - 1,473

Text Arena · Chinese · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena chinese leaderboard.

Rank #25 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,524
Percentile: 93.2%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: chinese. Source rank: #27. Votes: 1328. Organization: zai. License: MIT.

93.2% percentile inside its fair comparison set

1,524Raw benchmark valueCI 1,507 - 1,541

Text Arena · French · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena french leaderboard.

Rank #17 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,494
Percentile: 93.8%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: french. Source rank: #17. Votes: 532. Organization: zai. License: MIT.

93.8% percentile inside its fair comparison set

1,494Raw benchmark valueCI 1,468 - 1,520

Text Arena · German · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena german leaderboard.

Rank #5 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,504
Percentile: 98.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: german. Source rank: #5. Votes: 282. Organization: zai. License: MIT.

98.5% percentile inside its fair comparison set

1,504Raw benchmark valueCI 1,470 - 1,538

Text Arena · Japanese · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena japanese leaderboard.

Rank #30 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,445
Percentile: 88.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: japanese. Source rank: #33. Votes: 327. Organization: zai. License: MIT.

88.1% percentile inside its fair comparison set

1,445Raw benchmark valueCI 1,411 - 1,480

Text Arena · Korean · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena korean leaderboard.

Rank #7 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,475
Percentile: 97.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: korean. Source rank: #7. Votes: 298. Organization: zai. License: MIT.

97.5% percentile inside its fair comparison set

1,475Raw benchmark valueCI 1,441 - 1,510

Text Arena · Non English · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena non_english leaderboard.

Rank #31 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,457
Percentile: 92%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: non_english. Source rank: #33. Votes: 10629. Organization: zai. License: MIT.

92% percentile inside its fair comparison set

1,457Raw benchmark valueCI 1,450 - 1,463

Text Arena · Polish · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena polish leaderboard.

Rank #16 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,484
Percentile: 92.5%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: polish. Source rank: #17. Votes: 248. Organization: zai. License: MIT.

92.5% percentile inside its fair comparison set

1,484Raw benchmark valueCI 1,448 - 1,521

Text Arena · Russian · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena russian leaderboard.

Rank #38 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,458
Percentile: 89.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: russian. Source rank: #40. Votes: 2211. Organization: zai. License: MIT.

89.1% percentile inside its fair comparison set

1,458Raw benchmark valueCI 1,445 - 1,471

Text Arena · Spanish · No Style Control

AR · Multilingual · Human

Observed user preference in Arena's Text Arena spanish leaderboard.

Rank #63 · Source label: glm-5.3-max

verified runtimeexact alias

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,446
Percentile: 76.1%
Last updated: recent
Eligibility: headline eligible

Parsed from Arena leaderboard dataset row `glm-5.3-max`. Category: spanish. Source rank: #66. Votes: 538. Organization: zai. License: MIT.

76.1% percentile inside its fair comparison set

1,446Raw benchmark valueCI 1,421 - 1,470

Source links and registry checks

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Arena

Oct 3, 2026

source →

official

Artificial Analysis

Oct 3, 2026

source →

official

LiveBench

Oct 3, 2026

source →

official

Terminal-Bench

Oct 3, 2026

source →

glm-5.3-max

Current snapshot.

Source-linked scores by benchmark

Source links and registry checks

Loading model evidence.

glm-5.3-max

Current snapshot.

Source-linked scores by benchmark

Source links and registry checks