Model profile · DeepSeek

deepseek-v4-pro-thinking

Unknown weightsmid · registry tag 2026 benchmark-derived

Thin verified coverage

Reads as thin verified coverage across the resolved source data.

Visible coverage: 2.1%
Verified coverage: 2.1%
Spread: n/a
Last verified: Jun 20, 2026

textcode1 aliases31 official source links

Open compare

Data version

Current snapshot.

Data version Jun 20, 2026Model list checked9 providers · 1081 tracked modelsPage refreshed Jul 5, 2026

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

Source-linked scores by benchmark

Each row keeps the benchmark source, source type, raw metric, and percentile inside its fair comparison set.

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

Chat / text18 benchmarks91.8%

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 #26

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,458
Percentile: 92.3%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: overall. Source rank: #36. Votes: 26928. Organization: deepseek. License: MIT.

92.3% percentile inside its fair comparison set

1,458Raw benchmark valueCI 1,453 - 1,463

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 #25

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,442
Percentile: 92.6%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: creative_writing. Source rank: #34. Votes: 4315. Organization: deepseek. License: MIT.

92.6% percentile inside its fair comparison set

1,442Raw benchmark valueCI 1,432 - 1,452

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 #28

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,466
Percentile: 91.7%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: english. Source rank: #38. Votes: 12380. Organization: deepseek. License: MIT.

91.7% percentile inside its fair comparison set

1,466Raw benchmark valueCI 1,460 - 1,473

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 #27

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,460
Percentile: 92%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: exclude_ties. Source rank: #37. Votes: 20597. Organization: deepseek. License: MIT.

92% percentile inside its fair comparison set

1,460Raw benchmark valueCI 1,454 - 1,467

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 #29

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,478
Percentile: 91.4%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: hard_prompts. Source rank: #38. Votes: 17840. Organization: deepseek. License: MIT.

91.4% percentile inside its fair comparison set

1,478Raw benchmark valueCI 1,472 - 1,484

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 #29

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,484
Percentile: 91.4%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: hard_prompts_english. Source rank: #37. Votes: 8686. Organization: deepseek. License: MIT.

91.4% percentile inside its fair comparison set

1,484Raw benchmark valueCI 1,476 - 1,491

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 #27

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,450
Percentile: 92%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: instruction_following. Source rank: #35. Votes: 9073. Organization: deepseek. License: MIT.

92% percentile inside its fair comparison set

1,450Raw benchmark valueCI 1,442 - 1,457

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 #25

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,472
Percentile: 92.1%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: longer_query. Source rank: #33. Votes: 11456. Organization: deepseek. License: MIT.

92.1% percentile inside its fair comparison set

1,472Raw benchmark valueCI 1,465 - 1,479

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 #38

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,457
Percentile: 88.5%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: multi_turn. Source rank: #50. Votes: 4729. Organization: deepseek. License: MIT.

88.5% percentile inside its fair comparison set

1,457Raw benchmark valueCI 1,448 - 1,466

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 #26

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,447
Percentile: 92.3%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: overall. Source rank: #32. Votes: 26928. Organization: deepseek. License: MIT.

92.3% percentile inside its fair comparison set

1,447Raw benchmark valueCI 1,442 - 1,452

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 #18

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,442
Percentile: 94.7%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: creative_writing. Source rank: #23. Votes: 4315. Organization: deepseek. License: MIT.

94.7% percentile inside its fair comparison set

1,442Raw benchmark valueCI 1,432 - 1,452

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 #29

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

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

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: english. Source rank: #35. Votes: 12380. Organization: deepseek. License: MIT.

91.4% percentile inside its fair comparison set

1,454Raw benchmark valueCI 1,447 - 1,460

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 #27

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,443
Percentile: 92%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: exclude_ties. Source rank: #34. Votes: 20597. Organization: deepseek. License: MIT.

92% percentile inside its fair comparison set

1,443Raw benchmark valueCI 1,437 - 1,450

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 #25

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,456
Percentile: 92.6%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: hard_prompts. Source rank: #32. Votes: 17840. Organization: deepseek. License: MIT.

92.6% percentile inside its fair comparison set

1,456Raw benchmark valueCI 1,450 - 1,462

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 #24

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,460
Percentile: 92.9%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: hard_prompts_english. Source rank: #31. Votes: 8686. Organization: deepseek. License: MIT.

92.9% percentile inside its fair comparison set

1,460Raw benchmark valueCI 1,453 - 1,467

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 #28

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,436
Percentile: 91.7%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: instruction_following. Source rank: #35. Votes: 9073. Organization: deepseek. License: MIT.

91.7% percentile inside its fair comparison set

1,436Raw benchmark valueCI 1,429 - 1,443

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 #24

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,453
Percentile: 92.4%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: longer_query. Source rank: #31. Votes: 11456. Organization: deepseek. License: MIT.

92.4% percentile inside its fair comparison set

1,453Raw benchmark valueCI 1,446 - 1,459

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 #37

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,446
Percentile: 88.9%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: multi_turn. Source rank: #47. Votes: 4729. Organization: deepseek. License: MIT.

88.9% percentile inside its fair comparison set

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

Coding6 benchmarks78.7%

Code 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

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,458
Percentile: 75.3%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: overall. Source rank: #24. Votes: 6256. Organization: deepseek. License: MIT.

75.3% percentile inside its fair comparison set

1,458Raw benchmark valueCI 1,450 - 1,467

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

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,458
Percentile: 75.3%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: webdev. Source rank: #24. Votes: 6256. Organization: deepseek. License: MIT.

75.3% percentile inside its fair comparison set

1,458Raw benchmark valueCI 1,450 - 1,467

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 #18

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,464
Percentile: 76.7%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: webdev-html. Source rank: #24. Votes: 794. Organization: deepseek. License: MIT.

76.7% percentile inside its fair comparison set

1,464Raw benchmark valueCI 1,443 - 1,486

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 #19

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,452
Percentile: 69.5%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: webdev-react. Source rank: #24. Votes: 5447. Organization: deepseek. License: MIT.

69.5% percentile inside its fair comparison set

1,452Raw benchmark valueCI 1,443 - 1,461

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 #41

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,494
Percentile: 87.5%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: coding. Source rank: #51. Votes: 7788. Organization: deepseek. License: MIT.

87.5% percentile inside its fair comparison set

1,494Raw benchmark valueCI 1,486 - 1,501

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 #40

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,457
Percentile: 87.8%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: coding. Source rank: #50. Votes: 7788. Organization: deepseek. License: MIT.

87.8% percentile inside its fair comparison set

1,457Raw benchmark valueCI 1,450 - 1,465

Reasoning / math / science2 benchmarks95.1%

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 #14

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,477
Percentile: 95.9%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: math. Source rank: #18. Votes: 1391. Organization: deepseek. License: MIT.

95.9% percentile inside its fair comparison set

1,477Raw benchmark valueCI 1,461 - 1,493

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

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,465
Percentile: 94.3%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: math. Source rank: #23. Votes: 1391. Organization: deepseek. License: MIT.

94.3% percentile inside its fair comparison set

1,465Raw benchmark valueCI 1,449 - 1,482

Professional reasoning18 benchmarks91.4%

Text Arena · Expert

AR · Professional reasoning · Human

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

Rank #39

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,477
Percentile: 86.2%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: expert. Source rank: #47. Votes: 2341. Organization: deepseek. License: MIT.

86.2% percentile inside its fair comparison set

1,477Raw benchmark valueCI 1,464 - 1,490

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

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,455
Percentile: 90.9%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: industry_business_and_management_and_financial_operations. Source rank: #41. Votes: 5406. Organization: deepseek. License: MIT.

90.9% percentile inside its fair comparison set

1,455Raw benchmark valueCI 1,446 - 1,464

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 #31

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,426
Percentile: 90.7%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: industry_entertainment_and_sports_and_media. Source rank: #43. Votes: 5745. Organization: deepseek. License: MIT.

90.7% percentile inside its fair comparison set

1,426Raw benchmark valueCI 1,418 - 1,435

Text Arena · Industry Legal And Government

AR · Professional reasoning · Human

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

Rank #14

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,485
Percentile: 95.6%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: industry_legal_and_government. Source rank: #16. Votes: 2016. Organization: deepseek. License: MIT.

95.6% percentile inside its fair comparison set

1,485Raw benchmark valueCI 1,471 - 1,499

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 #29

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,478
Percentile: 91.3%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: industry_life_and_physical_and_social_science. Source rank: #38. Votes: 4278. Organization: deepseek. License: MIT.

91.3% percentile inside its fair comparison set

1,478Raw benchmark valueCI 1,468 - 1,488

Text Arena · Industry Mathematical

AR · Professional reasoning · Human

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

Rank #22

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,475
Percentile: 93.2%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: industry_mathematical. Source rank: #25. Votes: 1413. Organization: deepseek. License: MIT.

93.2% percentile inside its fair comparison set

1,475Raw benchmark valueCI 1,458 - 1,491

Text Arena · Industry Medicine And Healthcare

AR · Professional reasoning · Human

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

Rank #20

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,487
Percentile: 93.6%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: industry_medicine_and_healthcare. Source rank: #22. Votes: 1914. Organization: deepseek. License: MIT.

93.6% percentile inside its fair comparison set

1,487Raw benchmark valueCI 1,473 - 1,501

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 #39

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,487
Percentile: 88.3%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: industry_software_and_it_services. Source rank: #50. Votes: 10904. Organization: deepseek. License: MIT.

88.3% percentile inside its fair comparison set

1,487Raw benchmark valueCI 1,480 - 1,493

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 #30

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,444
Percentile: 91%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: industry_writing_and_literature_and_language. Source rank: #40. Votes: 6533. Organization: deepseek. License: MIT.

91% percentile inside its fair comparison set

1,444Raw benchmark valueCI 1,436 - 1,453

Text Arena · Expert · No Style Control

AR · Professional reasoning · Human

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

Rank #38

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,450
Percentile: 86.5%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: expert. Source rank: #46. Votes: 2341. Organization: deepseek. License: MIT.

86.5% percentile inside its fair comparison set

1,450Raw benchmark valueCI 1,437 - 1,463

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

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,434
Percentile: 90.9%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: industry_business_and_management_and_financial_operations. Source rank: #35. Votes: 5406. Organization: deepseek. License: MIT.

90.9% percentile inside its fair comparison set

1,434Raw benchmark valueCI 1,425 - 1,443

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 #27

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,421
Percentile: 92%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: industry_entertainment_and_sports_and_media. Source rank: #35. Votes: 5745. Organization: deepseek. License: MIT.

92% percentile inside its fair comparison set

1,421Raw benchmark valueCI 1,413 - 1,430

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 #18

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,470
Percentile: 94.3%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: industry_legal_and_government. Source rank: #23. Votes: 2016. Organization: deepseek. License: MIT.

94.3% percentile inside its fair comparison set

1,470Raw benchmark valueCI 1,457 - 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 #27

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,460
Percentile: 92%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: industry_life_and_physical_and_social_science. Source rank: #32. Votes: 4278. Organization: deepseek. License: MIT.

92% percentile inside its fair comparison set

1,460Raw benchmark valueCI 1,451 - 1,470

Text Arena · Industry Mathematical · No Style Control

AR · Professional reasoning · Human

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

Rank #26

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,460
Percentile: 91.9%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: industry_mathematical. Source rank: #31. Votes: 1413. Organization: deepseek. License: MIT.

91.9% percentile inside its fair comparison set

1,460Raw benchmark valueCI 1,443 - 1,476

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 #15

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,469
Percentile: 95.3%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: industry_medicine_and_healthcare. Source rank: #18. Votes: 1914. Organization: deepseek. License: MIT.

95.3% percentile inside its fair comparison set

1,469Raw benchmark valueCI 1,455 - 1,483

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 #38

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,459
Percentile: 88.6%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: industry_software_and_it_services. Source rank: #47. Votes: 10904. Organization: deepseek. License: MIT.

88.6% percentile inside its fair comparison set

1,459Raw benchmark valueCI 1,453 - 1,466

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 #24

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,438
Percentile: 92.9%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: industry_writing_and_literature_and_language. Source rank: #30. Votes: 6533. Organization: deepseek. License: MIT.

92.9% percentile inside its fair comparison set

1,438Raw benchmark valueCI 1,430 - 1,446

Multilingual14 benchmarks87.9%

Text Arena · Chinese

AR · Multilingual · Human

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

Rank #22

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,504
Percentile: 92.9%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: chinese. Source rank: #27. Votes: 1381. Organization: deepseek. License: MIT.

92.9% percentile inside its fair comparison set

1,504Raw benchmark valueCI 1,486 - 1,521

Text Arena · French

AR · Multilingual · Human

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

Rank #49

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,458
Percentile: 77.8%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: french. Source rank: #64. Votes: 988. Organization: deepseek. License: MIT.

77.8% percentile inside its fair comparison set

1,458Raw benchmark valueCI 1,437 - 1,479

Text Arena · German

AR · Multilingual · Human

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

Rank #27

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,454
Percentile: 89%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: german. Source rank: #34. Votes: 422. Organization: deepseek. License: MIT.

89% percentile inside its fair comparison set

1,454Raw benchmark valueCI 1,424 - 1,483

Text Arena · Japanese

AR · Multilingual · Human

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

Rank #17

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,441
Percentile: 92.1%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: japanese. Source rank: #22. Votes: 240. Organization: deepseek. License: MIT.

92.1% percentile inside its fair comparison set

1,441Raw benchmark valueCI 1,402 - 1,480

Text Arena · Korean

AR · Multilingual · Human

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

Rank #20

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,426
Percentile: 90.9%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: korean. Source rank: #26. Votes: 419. Organization: deepseek. License: MIT.

90.9% percentile inside its fair comparison set

1,426Raw benchmark valueCI 1,396 - 1,456

Text Arena · Russian

AR · Multilingual · Human

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

Rank #23

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,462
Percentile: 92.4%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: russian. Source rank: #32. Votes: 3034. Organization: deepseek. License: MIT.

92.4% percentile inside its fair comparison set

1,462Raw benchmark valueCI 1,450 - 1,473

Text Arena · Spanish

AR · Multilingual · Human

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

Rank #37

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,445
Percentile: 83.2%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: spanish. Source rank: #46. Votes: 821. Organization: deepseek. License: MIT.

83.2% percentile inside its fair comparison set

1,445Raw benchmark valueCI 1,423 - 1,467

Text Arena · Chinese · No Style Control

AR · Multilingual · Human

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

Rank #28

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,488
Percentile: 90.8%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: chinese. Source rank: #33. Votes: 1381. Organization: deepseek. License: MIT.

90.8% percentile inside its fair comparison set

1,488Raw benchmark valueCI 1,471 - 1,505

Text Arena · French · No Style Control

AR · Multilingual · Human

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

Rank #58

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,437
Percentile: 73.6%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: french. Source rank: #70. Votes: 988. Organization: deepseek. License: MIT.

73.6% percentile inside its fair comparison set

1,437Raw benchmark valueCI 1,416 - 1,458

Text Arena · German · No Style Control

AR · Multilingual · Human

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

Rank #25

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,442
Percentile: 89.9%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: german. Source rank: #30. Votes: 422. Organization: deepseek. License: MIT.

89.9% percentile inside its fair comparison set

1,442Raw benchmark valueCI 1,413 - 1,472

Text Arena · Japanese · No Style Control

AR · Multilingual · Human

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

Rank #18

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,424
Percentile: 91.6%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: japanese. Source rank: #22. Votes: 240. Organization: deepseek. License: MIT.

91.6% percentile inside its fair comparison set

1,424Raw benchmark valueCI 1,384 - 1,463

Text Arena · Korean · No Style Control

AR · Multilingual · Human

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

Rank #20

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,417
Percentile: 90.9%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: korean. Source rank: #25. Votes: 419. Organization: deepseek. License: MIT.

90.9% percentile inside its fair comparison set

1,417Raw benchmark valueCI 1,388 - 1,447

Text Arena · Russian · No Style Control

AR · Multilingual · Human

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

Rank #20

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,453
Percentile: 93.4%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: russian. Source rank: #27. Votes: 3034. Organization: deepseek. License: MIT.

93.4% percentile inside its fair comparison set

1,453Raw benchmark valueCI 1,441 - 1,464

Text Arena · Spanish · No Style Control

AR · Multilingual · Human

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

Rank #40

verified runtimeexact aliasBackground only

Raw row drilldownsource row, percentile, last updated, eligibility

Source: Arena
Raw value: 1,440
Percentile: 81.8%
Last updated: recent
Eligibility: benchmark_derived_model

Parsed from Arena leaderboard dataset row `deepseek-v4-pro-thinking`. Category: spanish. Source rank: #48. Votes: 821. Organization: deepseek. License: MIT.

81.8% percentile inside its fair comparison set

1,440Raw benchmark valueCI 1,418 - 1,462

Source links and registry checks

official

Arena

Jun 20, 2026

deepseek-v4-pro-thinking

Current snapshot.

Source-linked scores by benchmark

Source links and registry checks

Loading model evidence.

deepseek-v4-pro-thinking

Current snapshot.

Source-linked scores by benchmark

Source links and registry checks