Benchmarks · /benchmarks/scale-prbench-legal

PRBench Legal

Name: PRBench Legal
Creator: Scale Labs

Scale professional reasoning benchmark for legal analysis and review work.

Source · Scale Labs
Version · scale-labs snapshot 2026-10-03
Scores · 42

Test details

Visible tradeoffsThis is a rubric-judged signal, so it is more structured than arena taste but still depends on the scoring rubric.

source

Scale Labs

metric

Score (%)

judge

Rubric

direction

higher better

group id

scale_prbench_legal_current

domain

Professional reasoning

What it measures vs what it misses

✓ Measures

Applied legal reasoning on professional-domain tasks.

✗ Misses

Broad chat preference. Search freshness or real-time retrieval quality.

Why this countsApplied legal reasoning on professional-domain tasks.Same-test ruleThis percentile only compares models inside the exact benchmark/version group shown here. It is not a universal score.What it missesBroad chat preference.

Leaderboard · this benchmark version

#1 · muse-spark-1.3-max

SL · Oct 3, 2026

Source label: Muse Spark 1.3

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 100%
Last updated: recent
Eligibility: benchmark_derived_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: Muse Spark 1.3. Collapse policy: highest reported score per canonical model.

61.6%

#2 · muse-spark-1.1

SL · Oct 3, 2026

Source label: Muse Spark 1.1

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 97.6%
Last updated: recent
Eligibility: benchmark_derived_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: Muse Spark 1.1. Collapse policy: highest reported score per canonical model.

57.1%

#3 · Claude Fable 5

SL · Oct 3, 2026

Source label: claude-fable 5

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 95.1%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: claude-fable 5. Collapse policy: highest reported score per canonical model.

52.6%

#4 · muse-spark

SL · Oct 3, 2026

Source label: Muse Spark

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 92.7%
Last updated: recent
Eligibility: benchmark_derived_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: Muse Spark. Collapse policy: highest reported score per canonical model.

52.3%

#5 · Claude Opus 4.6

SL · Oct 3, 2026

Source label: claude-opus-4-6 (Non-Thinking)

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 90.2%
Last updated: recent
Eligibility: historical_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: claude-opus-4-6 (Non-Thinking). Collapse policy: highest reported score per canonical model.

52.3%

#6 · Fable 5.1

SL · Oct 3, 2026

Source label: Fable 5.1

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 87.8%
Last updated: recent
Eligibility: benchmark_derived_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: Fable 5.1. Collapse policy: highest reported score per canonical model.

51.6%

#7 · GPT-5.6 Sol

SL · Oct 3, 2026

Source label: gpt-5.6-sol (max)

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 85.4%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-5.6-sol (max). Collapse policy: highest reported score per canonical model.

50.5%

#8 · GPT-5

SL · Oct 3, 2026

Source label: gpt-5-pro

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 82.9%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-5-pro. Collapse policy: highest reported score per canonical model.

49.9%

#9 · GPT-5.4 mini

SL · Oct 3, 2026

Source label: gpt-5-pro

backfilledproxy backfilledBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 82.9%
Last updated: recent
Eligibility: Fallback benchmark identity is visible for context but excluded from default ranking.
Identity: benchmark proxy (0.58)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-5-pro. Collapse policy: highest reported score per canonical model. Backfilled from GPT-5 via approved benchmark identity mapping map-gpt-5-4-mini-to-gpt-5.

49.9%

#10 · GPT-5.4 nano

SL · Oct 3, 2026

Source label: gpt-5-pro

backfilledproxy backfilledBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 82.9%
Last updated: recent
Eligibility: Fallback benchmark identity is visible for context but excluded from default ranking.
Identity: benchmark proxy (0.58)

49.9%

#11 · o3-pro

SL · Oct 3, 2026

verified runtimeexact direct

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 75.6%
Last updated: recent
Eligibility: headline eligible
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: o3-pro. Collapse policy: highest reported score per canonical model.

49.7%

#12 · GPT-5.1

SL · Oct 3, 2026

Source label: gpt-5.1-thinking

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 73.2%
Last updated: recent
Eligibility: historical_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-5.1-thinking. Collapse policy: highest reported score per canonical model.

49.3%

#13 · gemini-3.8-flash-high

SL · Oct 3, 2026

Source label: Gemini 3.8 Flash

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 70.7%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: Gemini 3.8 Flash. Collapse policy: highest reported score per canonical model.

48.9%

#14 · o3

SL · Oct 3, 2026

verified runtimeexact directBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 68.3%
Last updated: recent
Eligibility: historical_model
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: o3. Collapse policy: highest reported score per canonical model.

48.6%

#15 · gpt-6-astra-max

SL · Oct 3, 2026

Source label: GPT 6 Astra

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 65.9%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: GPT 6 Astra. Collapse policy: highest reported score per canonical model.

48.4%

#16 · grok-4.7-xhigh

SL · Oct 3, 2026

Source label: Grok 4.7 (xHigh)

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 63.4%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: Grok 4.7 (xHigh). Collapse policy: highest reported score per canonical model.

47.6%

#17 · Claude Opus 5.5 (Adaptive Reasoning, Max Effort, Default Fallback)

SL · Oct 3, 2026

verified runtimeexact direct

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 61%
Last updated: recent
Eligibility: headline eligible
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: claude-opus-5-5. Collapse policy: highest reported score per canonical model.

47.6%

#18 · gpt-5.2-pro-2025-12-11

SL · Oct 3, 2026

Source label: gpt-5.2-pro-2025-12-11

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 58.5%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-5.2-pro-2025-12-11. Collapse policy: highest reported score per canonical model.

45.4%

#19 · GPT-5.4

SL · Oct 3, 2026

Source label: gpt-5.4 (High)

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 56.1%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-5.4 (High). Collapse policy: highest reported score per canonical model.

44.4%

#20 · claude-opus-4-5-20251101-thinking

SL · Oct 3, 2026

verified runtimeexact direct

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 53.7%
Last updated: recent
Eligibility: headline eligible
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: claude-opus-4-5-20251101-thinking. Collapse policy: highest reported score per canonical model.

44.2%

#21 · Gemini 3.1 Pro

SL · Oct 3, 2026

Source label: gemini-3.1-pro

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 51.2%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gemini-3.1-pro. Collapse policy: highest reported score per canonical model.

44%

#22 · Kimi K2.5 (Reasoning)

SL · Oct 3, 2026

Source label: kimi-k2.5

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 48.8%
Last updated: recent
Eligibility: benchmark_derived_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: kimi-k2.5. Collapse policy: highest reported score per canonical model.

43.8%

#23 · gpt-6-luna-max

SL · Oct 3, 2026

verified runtimeexact direct

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 46.3%
Last updated: recent
Eligibility: headline eligible
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-6-luna. Collapse policy: highest reported score per canonical model.

42.9%

#24 · Gemini 2.5 Pro

SL · Oct 3, 2026

Source label: gemini-2.5-pro

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 43.9%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gemini-2.5-pro. Collapse policy: highest reported score per canonical model.

41.4%

#25 · Gemini 2.5 Flash

SL · Oct 3, 2026

Source label: gemini-2.5-flash

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 41.5%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gemini-2.5-flash. Collapse policy: highest reported score per canonical model.

41%

#26 · Kimi K2 Thinking

SL · Oct 3, 2026

verified runtimeexact directBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 39%
Last updated: recent
Eligibility: benchmark_derived_model
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: kimi-k2-thinking. Collapse policy: highest reported score per canonical model.

40.9%

#27 · Claude Sonnet 4.5

SL · Oct 3, 2026

Source label: claude-sonnet-4-5-20250929

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 36.6%
Last updated: recent
Eligibility: historical_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: claude-sonnet-4-5-20250929. Collapse policy: highest reported score per canonical model.

40.8%

#28 · Gemini 3 Pro Preview

SL · Oct 3, 2026

verified runtimeexact directBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 34.1%
Last updated: recent
Eligibility: preview_model
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gemini-3-pro-preview. Collapse policy: highest reported score per canonical model.

40.6%

#29 · GPT-OSS 120B

SL · Oct 3, 2026

verified runtimeexact directBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 31.7%
Last updated: recent
Eligibility: historical_model
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-oss-120b. Collapse policy: highest reported score per canonical model.

40.2%

#30 · mistral-medium-latest

SL · Oct 3, 2026

verified runtimeexact direct

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 29.3%
Last updated: recent
Eligibility: headline eligible
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: mistral-medium-latest. Collapse policy: highest reported score per canonical model.

39.6%

#31 · gpt-6-sol-max

SL · Oct 3, 2026

verified runtimeexact direct

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 26.8%
Last updated: recent
Eligibility: headline eligible
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-6-sol. Collapse policy: highest reported score per canonical model.

38.6%

#32 · qwen.qwen3-235b-a22b-2507-v1:0

SL · Oct 3, 2026

Source label: qwen.qwen3-235b-a22b-2507-v1:0

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 24.4%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: qwen.qwen3-235b-a22b-2507-v1:0. Collapse policy: highest reported score per canonical model.

38.3%

#33 · o4 mini

SL · Oct 3, 2026

verified runtimeexact directBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 22%
Last updated: recent
Eligibility: historical_model
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: o4-mini. Collapse policy: highest reported score per canonical model.

38.1%

#34 · deepseek-v3p1

SL · Oct 3, 2026

verified runtimeexact direct

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 19.5%
Last updated: recent
Eligibility: headline eligible
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: deepseek-v3p1. Collapse policy: highest reported score per canonical model.

37.6%

#35 · deepseek-r1-0528

SL · Oct 3, 2026

verified runtimeexact direct

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 17.1%
Last updated: recent
Eligibility: headline eligible
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: deepseek-r1-0528. Collapse policy: highest reported score per canonical model.

36.6%

#36 · GPT-4.1

SL · Oct 3, 2026

Source label: gpt-4.1

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 14.6%
Last updated: recent
Eligibility: historical_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-4.1. Collapse policy: highest reported score per canonical model.

36.5%

#37 · kimi-k2-instruct

SL · Oct 3, 2026

verified runtimeexact directBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 12.2%
Last updated: recent
Eligibility: benchmark_derived_model
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: kimi-k2-instruct. Collapse policy: highest reported score per canonical model.

36.4%

#38 · Claude Opus 4.1

SL · Oct 3, 2026

Source label: claude-opus-4-1-20250805

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 9.8%
Last updated: recent
Eligibility: historical_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: claude-opus-4-1-20250805. Collapse policy: highest reported score per canonical model.

34%

#39 · Claude Opus 4

SL · Oct 3, 2026

Source label: claude-opus-4-1-20250805

backfilledproxy backfilledBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 9.8%
Last updated: recent
Eligibility: Fallback benchmark identity is visible for context but excluded from default ranking.
Identity: benchmark proxy (0.58)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: claude-opus-4-1-20250805. Collapse policy: highest reported score per canonical model. Backfilled from Claude Opus 4.1 via approved benchmark identity mapping map-claude-opus-4-to-4-1.

34%

#40 · Claude Opus 4.7

SL · Oct 3, 2026

Source label: claude-opus-4-1-20250805

backfilledproxy backfilledBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 9.8%
Last updated: recent
Eligibility: Fallback benchmark identity is visible for context but excluded from default ranking.
Identity: benchmark proxy (0.58)

34%

#41 · GPT-4.1 mini

SL · Oct 3, 2026

Source label: gpt-4.1-mini

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 2.4%
Last updated: recent
Eligibility: historical_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-4.1-mini. Collapse policy: highest reported score per canonical model.

30.4%

#42 · llama4-maverick-instruct-basic

SL · Oct 3, 2026

verified runtimeexact directBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 0%
Last updated: recent
Eligibility: benchmark_derived_model
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: llama4-maverick-instruct-basic. Collapse policy: highest reported score per canonical model.

24.8%

Benchmarks · /benchmarks/scale-prbench-legal

PRBench Legal

Scale professional reasoning benchmark for legal analysis and review work.

Source · Scale Labs
Version · scale-labs snapshot 2026-10-03
Scores · 42

Test details

Visible tradeoffsThis is a rubric-judged signal, so it is more structured than arena taste but still depends on the scoring rubric.

source

Scale Labs

metric

Score (%)

judge

Rubric

direction

higher better

group id

scale_prbench_legal_current

domain

Professional reasoning

What it measures vs what it misses

✓ Measures

Applied legal reasoning on professional-domain tasks.

✗ Misses

Broad chat preference. Search freshness or real-time retrieval quality.

Leaderboard · this benchmark version

#1 · muse-spark-1.3-max

SL · Oct 3, 2026

Source label: Muse Spark 1.3

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 100%
Last updated: recent
Eligibility: benchmark_derived_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: Muse Spark 1.3. Collapse policy: highest reported score per canonical model.

61.6%

#2 · muse-spark-1.1

SL · Oct 3, 2026

Source label: Muse Spark 1.1

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 97.6%
Last updated: recent
Eligibility: benchmark_derived_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: Muse Spark 1.1. Collapse policy: highest reported score per canonical model.

57.1%

#3 · Claude Fable 5

SL · Oct 3, 2026

Source label: claude-fable 5

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 95.1%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: claude-fable 5. Collapse policy: highest reported score per canonical model.

52.6%

#4 · muse-spark

SL · Oct 3, 2026

Source label: Muse Spark

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 92.7%
Last updated: recent
Eligibility: benchmark_derived_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: Muse Spark. Collapse policy: highest reported score per canonical model.

52.3%

#5 · Claude Opus 4.6

SL · Oct 3, 2026

Source label: claude-opus-4-6 (Non-Thinking)

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 90.2%
Last updated: recent
Eligibility: historical_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: claude-opus-4-6 (Non-Thinking). Collapse policy: highest reported score per canonical model.

52.3%

#6 · Fable 5.1

SL · Oct 3, 2026

Source label: Fable 5.1

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 87.8%
Last updated: recent
Eligibility: benchmark_derived_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: Fable 5.1. Collapse policy: highest reported score per canonical model.

51.6%

#7 · GPT-5.6 Sol

SL · Oct 3, 2026

Source label: gpt-5.6-sol (max)

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 85.4%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-5.6-sol (max). Collapse policy: highest reported score per canonical model.

50.5%

#8 · GPT-5

SL · Oct 3, 2026

Source label: gpt-5-pro

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 82.9%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-5-pro. Collapse policy: highest reported score per canonical model.

49.9%

#9 · GPT-5.4 mini

SL · Oct 3, 2026

Source label: gpt-5-pro

backfilledproxy backfilledBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 82.9%
Last updated: recent
Eligibility: Fallback benchmark identity is visible for context but excluded from default ranking.
Identity: benchmark proxy (0.58)

49.9%

#10 · GPT-5.4 nano

SL · Oct 3, 2026

Source label: gpt-5-pro

backfilledproxy backfilledBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 82.9%
Last updated: recent
Eligibility: Fallback benchmark identity is visible for context but excluded from default ranking.
Identity: benchmark proxy (0.58)

49.9%

#11 · o3-pro

SL · Oct 3, 2026

verified runtimeexact direct

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 75.6%
Last updated: recent
Eligibility: headline eligible
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: o3-pro. Collapse policy: highest reported score per canonical model.

49.7%

#12 · GPT-5.1

SL · Oct 3, 2026

Source label: gpt-5.1-thinking

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 73.2%
Last updated: recent
Eligibility: historical_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-5.1-thinking. Collapse policy: highest reported score per canonical model.

49.3%

#13 · gemini-3.8-flash-high

SL · Oct 3, 2026

Source label: Gemini 3.8 Flash

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 70.7%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: Gemini 3.8 Flash. Collapse policy: highest reported score per canonical model.

48.9%

#14 · o3

SL · Oct 3, 2026

verified runtimeexact directBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 68.3%
Last updated: recent
Eligibility: historical_model
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: o3. Collapse policy: highest reported score per canonical model.

48.6%

#15 · gpt-6-astra-max

SL · Oct 3, 2026

Source label: GPT 6 Astra

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 65.9%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: GPT 6 Astra. Collapse policy: highest reported score per canonical model.

48.4%

#16 · grok-4.7-xhigh

SL · Oct 3, 2026

Source label: Grok 4.7 (xHigh)

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 63.4%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: Grok 4.7 (xHigh). Collapse policy: highest reported score per canonical model.

47.6%

#17 · Claude Opus 5.5 (Adaptive Reasoning, Max Effort, Default Fallback)

SL · Oct 3, 2026

verified runtimeexact direct

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 61%
Last updated: recent
Eligibility: headline eligible
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: claude-opus-5-5. Collapse policy: highest reported score per canonical model.

47.6%

#18 · gpt-5.2-pro-2025-12-11

SL · Oct 3, 2026

Source label: gpt-5.2-pro-2025-12-11

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 58.5%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-5.2-pro-2025-12-11. Collapse policy: highest reported score per canonical model.

45.4%

#19 · GPT-5.4

SL · Oct 3, 2026

Source label: gpt-5.4 (High)

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 56.1%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-5.4 (High). Collapse policy: highest reported score per canonical model.

44.4%

#20 · claude-opus-4-5-20251101-thinking

SL · Oct 3, 2026

verified runtimeexact direct

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 53.7%
Last updated: recent
Eligibility: headline eligible
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: claude-opus-4-5-20251101-thinking. Collapse policy: highest reported score per canonical model.

44.2%

#21 · Gemini 3.1 Pro

SL · Oct 3, 2026

Source label: gemini-3.1-pro

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 51.2%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gemini-3.1-pro. Collapse policy: highest reported score per canonical model.

44%

#22 · Kimi K2.5 (Reasoning)

SL · Oct 3, 2026

Source label: kimi-k2.5

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 48.8%
Last updated: recent
Eligibility: benchmark_derived_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: kimi-k2.5. Collapse policy: highest reported score per canonical model.

43.8%

#23 · gpt-6-luna-max

SL · Oct 3, 2026

verified runtimeexact direct

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 46.3%
Last updated: recent
Eligibility: headline eligible
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-6-luna. Collapse policy: highest reported score per canonical model.

42.9%

#24 · Gemini 2.5 Pro

SL · Oct 3, 2026

Source label: gemini-2.5-pro

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 43.9%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gemini-2.5-pro. Collapse policy: highest reported score per canonical model.

41.4%

#25 · Gemini 2.5 Flash

SL · Oct 3, 2026

Source label: gemini-2.5-flash

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 41.5%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gemini-2.5-flash. Collapse policy: highest reported score per canonical model.

41%

#26 · Kimi K2 Thinking

SL · Oct 3, 2026

verified runtimeexact directBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 39%
Last updated: recent
Eligibility: benchmark_derived_model
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: kimi-k2-thinking. Collapse policy: highest reported score per canonical model.

40.9%

#27 · Claude Sonnet 4.5

SL · Oct 3, 2026

Source label: claude-sonnet-4-5-20250929

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 36.6%
Last updated: recent
Eligibility: historical_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: claude-sonnet-4-5-20250929. Collapse policy: highest reported score per canonical model.

40.8%

#28 · Gemini 3 Pro Preview

SL · Oct 3, 2026

verified runtimeexact directBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 34.1%
Last updated: recent
Eligibility: preview_model
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gemini-3-pro-preview. Collapse policy: highest reported score per canonical model.

40.6%

#29 · GPT-OSS 120B

SL · Oct 3, 2026

verified runtimeexact directBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 31.7%
Last updated: recent
Eligibility: historical_model
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-oss-120b. Collapse policy: highest reported score per canonical model.

40.2%

#30 · mistral-medium-latest

SL · Oct 3, 2026

verified runtimeexact direct

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 29.3%
Last updated: recent
Eligibility: headline eligible
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: mistral-medium-latest. Collapse policy: highest reported score per canonical model.

39.6%

#31 · gpt-6-sol-max

SL · Oct 3, 2026

verified runtimeexact direct

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 26.8%
Last updated: recent
Eligibility: headline eligible
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-6-sol. Collapse policy: highest reported score per canonical model.

38.6%

#32 · qwen.qwen3-235b-a22b-2507-v1:0

SL · Oct 3, 2026

Source label: qwen.qwen3-235b-a22b-2507-v1:0

verified runtimeexact alias

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 24.4%
Last updated: recent
Eligibility: headline eligible
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: qwen.qwen3-235b-a22b-2507-v1:0. Collapse policy: highest reported score per canonical model.

38.3%

#33 · o4 mini

SL · Oct 3, 2026

verified runtimeexact directBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 22%
Last updated: recent
Eligibility: historical_model
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: o4-mini. Collapse policy: highest reported score per canonical model.

38.1%

#34 · deepseek-v3p1

SL · Oct 3, 2026

verified runtimeexact direct

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 19.5%
Last updated: recent
Eligibility: headline eligible
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: deepseek-v3p1. Collapse policy: highest reported score per canonical model.

37.6%

#35 · deepseek-r1-0528

SL · Oct 3, 2026

verified runtimeexact direct

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 17.1%
Last updated: recent
Eligibility: headline eligible
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: deepseek-r1-0528. Collapse policy: highest reported score per canonical model.

36.6%

#36 · GPT-4.1

SL · Oct 3, 2026

Source label: gpt-4.1

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 14.6%
Last updated: recent
Eligibility: historical_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-4.1. Collapse policy: highest reported score per canonical model.

36.5%

#37 · kimi-k2-instruct

SL · Oct 3, 2026

verified runtimeexact directBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 12.2%
Last updated: recent
Eligibility: benchmark_derived_model
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: kimi-k2-instruct. Collapse policy: highest reported score per canonical model.

36.4%

#38 · Claude Opus 4.1

SL · Oct 3, 2026

Source label: claude-opus-4-1-20250805

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 9.8%
Last updated: recent
Eligibility: historical_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: claude-opus-4-1-20250805. Collapse policy: highest reported score per canonical model.

34%

#39 · Claude Opus 4

SL · Oct 3, 2026

Source label: claude-opus-4-1-20250805

backfilledproxy backfilledBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 9.8%
Last updated: recent
Eligibility: Fallback benchmark identity is visible for context but excluded from default ranking.
Identity: benchmark proxy (0.58)

34%

#40 · Claude Opus 4.7

SL · Oct 3, 2026

Source label: claude-opus-4-1-20250805

backfilledproxy backfilledBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 9.8%
Last updated: recent
Eligibility: Fallback benchmark identity is visible for context but excluded from default ranking.
Identity: benchmark proxy (0.58)

34%

#41 · GPT-4.1 mini

SL · Oct 3, 2026

Source label: gpt-4.1-mini

verified runtimeexact aliasBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 2.4%
Last updated: recent
Eligibility: historical_model
Identity: provider alias (0.92)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: gpt-4.1-mini. Collapse policy: highest reported score per canonical model.

30.4%

#42 · llama4-maverick-instruct-basic

SL · Oct 3, 2026

verified runtimeexact directBackground only

Raw row drilldownsource, percentile, eligibility

Source URL: https://labs.scale.com/leaderboard/prbench-legal
Percentile: 0%
Last updated: recent
Eligibility: benchmark_derived_model
Identity: exact (1.00)

Parsed from the public Scale Labs page for scale-prbench-legal. Reported model configuration: llama4-maverick-instruct-basic. Collapse policy: highest reported score per canonical model.

24.8%

PRBench Legal

Test details

What it measures vs what it misses

✓ Measures

✗ Misses

Leaderboard · this benchmark version

Loading benchmark evidence.

PRBench Legal

Test details

What it measures vs what it misses

✓ Measures

✗ Misses

Leaderboard · this benchmark version