# LegalBench / issue tasks > Source: Benchize (https://benchize.com) — https://benchize.com/benchmarks/vals-ai%3Alegal_bench%3Aissue_tasks. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/benchmarks/vals-ai%3Alegal_bench%3Aissue_tasks. Evaluating language models on a wide range of open source legal reasoning tasks. Canonical URL: https://benchize.com/benchmarks/vals-ai%3Alegal_bench%3Aissue_tasks Last updated: Wed, 21 Oct 2026 18:52:20 GMT Reproduction and quotation must name Benchize, link to the canonical HTML URL, and retain individual source citations. Evaluating language models on a wide range of open source legal reasoning tasks. Professional work · Models tested: 44 · Score publishers: Vals AI ## Model results | # | Model | Score | ± | Cost | Latency | | | --- | --- | --- | --- | --- | --- | --- | | 1 | [Gemini 4 Argon](https://benchize.com/models/gemini-4-argon-high.md)High | 92.94% | SE ±0.5 pp | $0.01 | 8.44source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 92.941 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.014298 \| USD \| per_test \| Lower is better \| \| `latency` \| 8.441 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 262144 provider Google reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.5 pp sourcePopulation 151 stderr 0.497 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label google/gemini-4-argon Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/google~1gemini-4-argon` | | | | | | | | 2 | [Claude Fable 5.1](https://benchize.com/models/claude-fable-5-1-max.md)Max | 92.72% | SE ±0.5 pp | $0.04 | 6.13source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 92.72 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.037363 \| USD \| per_test \| Lower is better \| \| `latency` \| 6.13 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort max harness Not reported maxOutputTokens 128000 provider Anthropic reasoning Not reported Reasoning effort max Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.5 pp sourcePopulation 151 stderr 0.498 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label anthropic/claude-fable-5-1 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/anthropic~1claude-fable-5-1` | | | | | | | | 3 | [Gemini 3.7 Flash](https://benchize.com/models/gemini-3-7-flash-high.md)High | 92.15% | SE ±0.51 pp | $0.00449 | 2.02source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 92.146 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.004487 \| USD \| per_test \| Lower is better \| \| `latency` \| 2.017 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 65536 provider Google reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.51 pp sourcePopulation 151 stderr 0.508 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label google/gemini-3.7-flash Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/google~1gemini-3.7-flash` | | | | | | | | 4 | [Claude Fable 5](https://benchize.com/models/claude-fable-5-max.md)Max | 92.06% | SE ±0.51 pp | $0.03 | 6.92source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 92.064 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.030005 \| USD \| per_test \| Lower is better \| \| `latency` \| 6.922 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort max harness Not reported maxOutputTokens 128000 provider Anthropic reasoning Not reported Reasoning effort max Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.51 pp sourcePopulation 151 stderr 0.505 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label anthropic/claude-fable-5 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/anthropic~1claude-fable-5` | | | | | | | | 5 | [Claude Opus 5](https://benchize.com/models/claude-opus-5-max.md)Max | 91.88% | SE ±0.53 pp | $0.02 | 2.89source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 91.878 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.015445 \| USD \| per_test \| Lower is better \| \| `latency` \| 2.886 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort max harness Not reported maxOutputTokens 128000 provider Anthropic reasoning Not reported Reasoning effort max Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.53 pp sourcePopulation 151 stderr 0.527 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label anthropic/claude-opus-5 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/anthropic~1claude-opus-5` | | | | | | | | 6 | [Gemini 3.1 Pro Preview](https://benchize.com/models/gemini-3-1-pro-preview-high.md)High | 91.48% | SE ±0.53 pp | $0.00681 | 8.72source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 91.48 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.006807 \| USD \| per_test \| Lower is better \| \| `latency` \| 8.723 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 65536 provider Google reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.53 pp sourcePopulation 151 stderr 0.53 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label google/gemini-3.1-pro-preview Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/google~1gemini-3.1-pro-preview` | | | | | | | | 7 | [GPT-5.6 Sol](https://benchize.com/models/gpt-5-6-sol-max.md)Max | 91.44% | SE ±0.54 pp | $0.00854 | 3source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 91.438 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.008536 \| USD \| per_test \| Lower is better \| \| `latency` \| 3.001 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 128000 provider OpenAI reasoning Not reported Reasoning effort max Temperature Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.54 pp sourcePopulation 151 stderr 0.535 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label openai/gpt-5.6-sol Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/openai~1gpt-5.6-sol` | | | | | | | | 8 | [Gemini 3.8 Flash](https://benchize.com/models/gemini-3-8-flash-high.md)High | 91.28% | SE ±0.53 pp | $0.00623 | 2.36source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 91.278 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.006227 \| USD \| per_test \| Lower is better \| \| `latency` \| 2.361 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 65536 provider Google reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.53 pp sourcePopulation 151 stderr 0.533 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label google/gemini-3.8-flash Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/google~1gemini-3.8-flash` | | | | | | | | 9 | [GPT-5.5](https://benchize.com/models/gpt-5-5-xhigh.md)Extra high | 91.19% | SE ±0.55 pp | $0.00728 | 6.8source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 91.193 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.00728 \| USD \| per_test \| Lower is better \| \| `latency` \| 6.799 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 128000 provider OpenAI reasoning Not reported Reasoning effort xhigh Temperature Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.55 pp sourcePopulation 151 stderr 0.547 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label openai/gpt-5.5 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/openai~1gpt-5.5` | | | | | | | | 10 | [Gemini 3 Flash Preview](https://benchize.com/models/gemini-3-flash-preview-high.md)High | 91.15% | SE ±0.54 pp | $0.0021 | 4.21source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 91.145 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.002102 \| USD \| per_test \| Lower is better \| \| `latency` \| 4.207 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 65536 provider Google reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.54 pp sourcePopulation 151 stderr 0.537 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label google/gemini-3-flash-preview Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/google~1gemini-3-flash-preview` | | | | | | | | 11 | [Gemini 3.6 Flash](https://benchize.com/models/gemini-3-6-flash-high.md)High | 91.12% | SE ±0.56 pp | $0.00511 | 2.52source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 91.121 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.005112 \| USD \| per_test \| Lower is better \| \| `latency` \| 2.519 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 65536 provider Google reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.56 pp sourcePopulation 151 stderr 0.557 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label google/gemini-3.6-flash Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/google~1gemini-3.6-flash` | | | | | | | | 12 | [MiniMax M3](https://benchize.com/models/minimax-m3.md) | 90.86% | SE ±0.55 pp | $0.0009 | 6.98source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 90.862 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.0009 \| USD \| per_test \| Lower is better \| \| `latency` \| 6.976 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 512000 provider MiniMax reasoning Not reported Reasoning effort unknown Temperature 1 topP 0.95 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.55 pp sourcePopulation 151 stderr 0.553 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label minimax/MiniMax-M3 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/minimax~1MiniMax-M3` | | | | | | | | 13 | [GPT-5.6 Terra](https://benchize.com/models/gpt-5-6-terra-max.md)Max | 89.39% | SE ±0.57 pp | $0.00284 | 2.16source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 89.392 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.002836 \| USD \| per_test \| Lower is better \| \| `latency` \| 2.159 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 128000 provider OpenAI reasoning Not reported Reasoning effort max Temperature Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.57 pp sourcePopulation 151 stderr 0.567 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label openai/gpt-5.6-terra Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/openai~1gpt-5.6-terra` | | | | | | | | 14 | [Muse Spark 1.2](https://benchize.com/models/muse-spark-1-2-xhigh.md)Extra high | 89.37% | SE ±0.58 pp | $0.00529 | 21.72source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 89.374 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.005294 \| USD \| per_test \| Lower is better \| \| `latency` \| 21.721 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 131072 provider Meta reasoning Not reported Reasoning effort xhigh Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.58 pp sourcePopulation 151 stderr 0.58 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label meta/muse_spark_1_2 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/meta~1muse_spark_1_2` | | | | | | | | 15 | [Gemini 3.1 Flash Lite Preview](https://benchize.com/models/gemini-3-1-flash-lite-preview-high.md)High | 89.37% | SE ±0.57 pp | $0.000329 | 2.39source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 89.371 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.000329 \| USD \| per_test \| Lower is better \| \| `latency` \| 2.392 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 65536 provider Google reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.57 pp sourcePopulation 151 stderr 0.57 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label google/gemini-3.1-flash-lite-preview Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/google~1gemini-3.1-flash-lite-preview` | | | | | | | | 16 | [Gemini 3.5 Flash-Lite](https://benchize.com/models/gemini-3-5-flash-lite-high.md)High | 89.33% | SE ±0.58 pp | $0.00119 | 1.33source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 89.325 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.001185 \| USD \| per_test \| Lower is better \| \| `latency` \| 1.33 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 65536 provider Google reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.58 pp sourcePopulation 151 stderr 0.582 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label google/gemini-3.5-flash-lite Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/google~1gemini-3.5-flash-lite` | | | | | | | | 17 | [Mercury 2.5](https://benchize.com/models/mercury-2-5-high.md)High | 89.05% | SE ±0.57 pp | $0.000911 | 3.59source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 89.051 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.000911 \| USD \| per_test \| Lower is better \| \| `latency` \| 3.586 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 65536 provider Inception reasoning Not reported Reasoning effort high Temperature Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.57 pp sourcePopulation 151 stderr 0.569 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label inception/mercury-2.5 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/inception~1mercury-2.5` | | | | | | | | 18 | [Muse Spark 1.1](https://benchize.com/models/muse-spark-1-1-xhigh.md)Extra high | 88.99% | SE ±0.57 pp | $0.00992 | 16.81source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 88.993 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.009923 \| USD \| per_test \| Lower is better \| \| `latency` \| 16.813 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 256000 provider Meta reasoning Not reported Reasoning effort xhigh Temperature Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.57 pp sourcePopulation 151 stderr 0.571 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label meta/muse_spark_1_1 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/meta~1muse_spark_1_1` | | | | | | | | 19 | [Claude Opus 4.7](https://benchize.com/models/claude-opus-4-7-max.md)Max | 88.95% | SE ±0.57 pp | $0.01 | 27.49source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 88.947 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.013182 \| USD \| per_test \| Lower is better \| \| `latency` \| 27.488 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort max harness Not reported maxOutputTokens 128000 provider Anthropic reasoning Not reported Reasoning effort max Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.57 pp sourcePopulation 151 stderr 0.573 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label anthropic/claude-opus-4-7 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/anthropic~1claude-opus-4-7` | | | | | | | | 20 | [DeepSeek V4.1 Flash](https://benchize.com/models/deepseek-v4-1-flash-high.md)High | 88.75% | SE ±0.6 pp | $0.000887 | 4.23source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 88.753 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.000887 \| USD \| per_test \| Lower is better \| \| `latency` \| 4.229 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 384000 provider DeepSeek reasoning Not reported Reasoning effort high Temperature Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.6 pp sourcePopulation 151 stderr 0.597 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label deepseek/deepseek-v4.1-flash Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/deepseek~1deepseek-v4.1-flash` | | | | | | | | 21 | [GPT-5.6 Luna](https://benchize.com/models/gpt-5-6-luna-max.md)Max | 88.68% | SE ±0.59 pp | $0.000489 | 2.82source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 88.68 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.000489 \| USD \| per_test \| Lower is better \| \| `latency` \| 2.819 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 128000 provider OpenAI reasoning Not reported Reasoning effort max Temperature Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.59 pp sourcePopulation 151 stderr 0.591 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label openai/gpt-5.6-luna Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/openai~1gpt-5.6-luna` | | | | | | | | 22 | [Claude Sonnet 5](https://benchize.com/models/claude-sonnet-5-max.md)Max | 88.24% | SE ±0.61 pp | $0.00574 | 2.2source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 88.24 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.00574 \| USD \| per_test \| Lower is better \| \| `latency` \| 2.196 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort max harness Not reported maxOutputTokens 128000 provider Anthropic reasoning Not reported Reasoning effort max Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.61 pp sourcePopulation 151 stderr 0.609 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label anthropic/claude-sonnet-5 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/anthropic~1claude-sonnet-5` | | | | | | | | 23 | [Claude Opus 4.8](https://benchize.com/models/claude-opus-4-8-max.md)Max | 87.83% | SE ±0.6 pp | $0.01 | 2.14source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 87.834 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.013321 \| USD \| per_test \| Lower is better \| \| `latency` \| 2.135 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort max harness Not reported maxOutputTokens 128000 provider Anthropic reasoning Not reported Reasoning effort max Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.6 pp sourcePopulation 151 stderr 0.601 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label anthropic/claude-opus-4-8 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/anthropic~1claude-opus-4-8` | | | | | | | | 24 | [GPT-4o (2024-11-20)](https://benchize.com/models/gpt-4o-2024-11-20-high.md)High | 87.57% | SE ±0.62 pp | — | 0.38source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 87.568 \| percent \| reported \| Higher is better \| \| `latency` \| 0.381 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 16384 provider OpenAI reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.62 pp sourcePopulation 151 stderr 0.615 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label openai/gpt-4o-2024-11-20 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/openai~1gpt-4o-2024-11-20` | | | | | | | | 25 | [Nemotron 3 Ultra](https://benchize.com/models/nemotron-3-ultra-550b-a55b.md) | 87.32% | SE ±0.59 pp | — | 2.07source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 87.318 \| percent \| reported \| Higher is better \| \| `latency` \| 2.071 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 128000 provider Nvidia reasoning Not reported Reasoning effort unknown Temperature Not reported topP 0.95 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.59 pp sourcePopulation 151 stderr 0.59 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label nvidia/nemotron-3-ultra-550b-a55b Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/nvidia~1nemotron-3-ultra-550b-a55b` | | | | | | | | 26 | [MiniMax M2.7](https://benchize.com/models/minimax-m2-7.md) | 87.12% | SE ±0.64 pp | $0.000927 | 8.84source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 87.118 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.000927 \| USD \| per_test \| Lower is better \| \| `latency` \| 8.837 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 196608 provider MiniMax reasoning Not reported Reasoning effort unknown Temperature 1 topP 0.95 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.64 pp sourcePopulation 151 stderr 0.64 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label minimax/MiniMax-M2.7 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/minimax~1MiniMax-M2.7` | | | | | | | | 27 | [Claude Sonnet 4.6](https://benchize.com/models/claude-sonnet-4-6-max.md)Max | 86.57% | SE ±0.59 pp | $0.00522 | 20.23source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 86.574 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.005224 \| USD \| per_test \| Lower is better \| \| `latency` \| 20.233 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort max harness Not reported maxOutputTokens 128000 provider Anthropic reasoning true Reasoning effort max Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.59 pp sourcePopulation 151 stderr 0.593 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label anthropic/claude-sonnet-4-6 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/anthropic~1claude-sonnet-4-6` | | | | | | | | 28 | [Gemini 2.5 Flash Lite Preview 09-2025](https://benchize.com/models/gemini-2-5-flash-lite-preview-09-2025.md) | 86.45% | SE ±0.62 pp | — | 0.4source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 86.447 \| percent \| reported \| Higher is better \| \| `latency` \| 0.402 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 65535 provider Google reasoning Not reported Reasoning effort unknown Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.62 pp sourcePopulation 151 stderr 0.619 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label google/gemini-2.5-flash-lite-preview-09-2025 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/google~1gemini-2.5-flash-lite-preview-09-2025` | | | | | | | | 29 | [MiMo-V2.5](https://benchize.com/models/mimo-v2-5.md) | 86.34% | SE ±0.63 pp | $0.000331 | 12.93source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 86.337 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.000331 \| USD \| per_test \| Lower is better \| \| `latency` \| 12.928 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 128000 provider Xiaomi reasoning Not reported Reasoning effort unknown Temperature 1 topP 0.95 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.63 pp sourcePopulation 151 stderr 0.626 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label xiaomi/mimo-v2.5 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/xiaomi~1mimo-v2.5` | | | | | | | | 30 | [DeepSeek V4 Pro 0813](https://benchize.com/models/deepseek-v4-pro-0813-max.md)Max | 85.99% | SE ±0.66 pp | $0.00226 | 8.55source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 85.989 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.002262 \| USD \| per_test \| Lower is better \| \| `latency` \| 8.552 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 384000 provider DeepSeek reasoning Not reported Reasoning effort max Temperature Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.66 pp sourcePopulation 151 stderr 0.661 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label deepseek/deepseek-v4-pro-0813 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/deepseek~1deepseek-v4-pro-0813` | | | | | | | | 31 | [Command A](https://benchize.com/models/command-a.md) | 84.8% | SE ±0.62 pp | — | 0.76source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 84.798 \| percent \| reported \| Higher is better \| \| `latency` \| 0.762 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 8000 provider Cohere reasoning Not reported Reasoning effort unknown Temperature 0.3 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.62 pp sourcePopulation 151 stderr 0.615 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label cohere/command-a-03-2025 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/cohere~1command-a-03-2025` | | | | | | | | 32 | [DeepSeek V4 Pro](https://benchize.com/models/deepseek-v4-pro-max.md)Max | 84.59% | SE ±0.67 pp | $0.00519 | 283.66source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 84.587 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.00519 \| USD \| per_test \| Lower is better \| \| `latency` \| 283.655 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 256000 provider DeepSeek reasoning Not reported Reasoning effort max Temperature Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.67 pp sourcePopulation 151 stderr 0.671 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label deepseek/deepseek-v4-pro Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/deepseek~1deepseek-v4-pro` | | | | | | | | 33 | [MiniMax M2.5](https://benchize.com/models/minimax-m2-5.md) | 84.17% | SE ±0.66 pp | $0.00131 | 14.9source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 84.169 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.001305 \| USD \| per_test \| Lower is better \| \| `latency` \| 14.901 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 196608 provider MiniMax reasoning Not reported Reasoning effort unknown Temperature 1 topP 0.95 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.66 pp sourcePopulation 151 stderr 0.662 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label minimax/MiniMax-M2.5 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/minimax~1MiniMax-M2.5` | | | | | | | | 34 | [GPT-4 Turbo](https://benchize.com/models/gpt-4-turbo-high.md)High | 82.88% | SE ±0.5 pp | — | 1.17source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 82.88 \| percent \| reported \| Higher is better \| \| `latency` \| 1.166 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 4096 provider OpenAI reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.5 pp sourcePopulation 151 stderr 0.499 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label openai/gpt-4-turbo Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/openai~1gpt-4-turbo` | | | | | | | | 35 | [DeepSeek V4 Flash 0731](https://benchize.com/models/deepseek-v4-flash-0731-high.md)High | 81.96% | SE ±0.73 pp | $0.000484 | 2.83source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 81.956 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.000484 \| USD \| per_test \| Lower is better \| \| `latency` \| 2.829 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 256000 provider DeepSeek reasoning Not reported Reasoning effort high Temperature Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.73 pp sourcePopulation 151 stderr 0.727 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label deepseek/deepseek-v4-flash-0731 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/deepseek~1deepseek-v4-flash-0731` | | | | | | | | 36 | [Mistral Large 3 2512](https://benchize.com/models/mistral-large-2512.md) | 81.01% | SE ±0.69 pp | $0.0008 | 0.94source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 81.006 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.0008 \| USD \| per_test \| Lower is better \| \| `latency` \| 0.938 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 256000 provider Mistral AI reasoning Not reported Reasoning effort unknown Temperature 0.065 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.69 pp sourcePopulation 151 stderr 0.694 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label mistralai/mistral-large-2512 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/mistralai~1mistral-large-2512` | | | | | | | | 37 | [GPT-4o (2024-08-06)](https://benchize.com/models/gpt-4o-2024-08-06-high.md)High | 80.6% | SE ±0.65 pp | — | 0.42source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 80.596 \| percent \| reported \| Higher is better \| \| `latency` \| 0.42 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 16384 provider OpenAI reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.65 pp sourcePopulation 151 stderr 0.653 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label openai/gpt-4o-2024-08-06 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/openai~1gpt-4o-2024-08-06` | | | | | | | | 38 | [MiMo-V2.5-Pro](https://benchize.com/models/mimo-v2-5-pro.md) | 80.17% | SE ±0.77 pp | $0.00105 | 32.35source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 80.173 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.00105 \| USD \| per_test \| Lower is better \| \| `latency` \| 32.35 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 128000 provider Xiaomi reasoning Not reported Reasoning effort unknown Temperature 1 topP 0.95 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.77 pp sourcePopulation 151 stderr 0.769 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label xiaomi/mimo-v2.5-pro Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/xiaomi~1mimo-v2.5-pro` | | | | | | | | 39 | [Laguna M 1](https://benchize.com/models/laguna-m-1.md) | 78.46% | SE ±0.79 pp | — | 29.01source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 78.459 \| percent \| reported \| Higher is better \| \| `latency` \| 29.012 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 32768 provider Poolside reasoning Not reported Reasoning effort unknown Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.79 pp sourcePopulation 151 stderr 0.791 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label poolside/laguna-m.1 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/poolside~1laguna-m.1` | | | | | | | | 40 | [Laguna Xs 2](https://benchize.com/models/laguna-xs-2.md) | 76.81% | SE ±0.77 pp | — | 9.39source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 76.811 \| percent \| reported \| Higher is better \| \| `latency` \| 9.388 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 32768 provider Poolside reasoning Not reported Reasoning effort unknown Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.77 pp sourcePopulation 151 stderr 0.771 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label poolside/laguna-xs.2 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/poolside~1laguna-xs.2` | | | | | | | | 41 | [Gemini 3.5 Flash](https://benchize.com/models/gemini-3-5-flash-high.md)High | 76.01% | SE ±3.49 pp | $0.00786 | 50.77source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 76.014 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.007857 \| USD \| per_test \| Lower is better \| \| `latency` \| 50.772 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 65536 provider Google reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±3.49 pp sourcePopulation 151 stderr 3.493 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label google/gemini-3.5-flash Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/google~1gemini-3.5-flash` | | | | | | | | 42 | [MiniMax M2.1](https://benchize.com/models/minimax-m2-1.md) | 71.97% | SE ±0.81 pp | $0.00105 | 10.12source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 71.973 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.001045 \| USD \| per_test \| Lower is better \| \| `latency` \| 10.116 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 131000 provider MiniMax reasoning Not reported Reasoning effort unknown Temperature 1 topP 0.95 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.81 pp sourcePopulation 151 stderr 0.807 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label minimax/MiniMax-M2.1 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/minimax~1MiniMax-M2.1` | | | | | | | | 43 | [GPT-3.5 Turbo](https://benchize.com/models/gpt-3-5-turbo-high.md)High | 57.05% | SE ±0.52 pp | — | 0.67source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 57.054 \| percent \| reported \| Higher is better \| \| `latency` \| 0.665 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 4096 provider OpenAI reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.52 pp sourcePopulation 151 stderr 0.515 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label openai/gpt-3.5-turbo Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/openai~1gpt-3.5-turbo` | | | | | | | | 44 | [Command A Plus](https://benchize.com/models/command-a-plus.md) | 55.18% | SE ±0.87 pp | $0.00896 | 5.36source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 55.176 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.008963 \| USD \| per_test \| Lower is better \| \| `latency` \| 5.359 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 64000 provider Cohere reasoning Not reported Reasoning effort unknown Temperature 0.9 topP 0.95 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.87 pp sourcePopulation 151 stderr 0.865 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_bench.D6uTgA3b.json) Source Vals AI Original model label cohere/command-a-plus-05-2026 Source version 1 Observed 2026-10-21T16:45:53.390Z Evidence SHA-256 `c8486ba756baa5cd3955d035edff67440e8db23e69d62ad1a9e18d15b8c4bc52` Source locator `/tasks/issue_tasks/cohere~1command-a-plus-05-2026` | | | | | | | --- Cite as: LegalBench / issue tasks — Benchize, https://benchize.com/benchmarks/vals-ai%3Alegal_bench%3Aissue_tasks (retrieved 2026-10-21)