# HLAB-Vals-1.1 / criteria pass rate white collar defense investigations > Source: Benchize (https://benchize.com) — https://benchize.com/benchmarks/vals-ai%3Ahlab%3Acriteria_pass_rate__white_collar_defense_investigations. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/benchmarks/vals-ai%3Ahlab%3Acriteria_pass_rate__white_collar_defense_investigations. Tests an agent's ability to complete legal work using documents, spreadsheets, presentations, and file-system tools Canonical URL: https://benchize.com/benchmarks/vals-ai%3Ahlab%3Acriteria_pass_rate__white_collar_defense_investigations 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. Tests an agent's ability to complete legal work using documents, spreadsheets, presentations, and file-system tools Professional work · Models tested: 14 · Score publishers: Vals AI ## Model results | # | Model | Score | ± | Cost | Latency | | | --- | --- | --- | --- | --- | --- | --- | | 1 | [Muse Wonder](https://benchize.com/models/muse-wonder-max.md)Max | 96.97% | SE ±1 pp | $6.12 | 1,201.64source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 96.97 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 6.121717 \| USD \| per_test \| Lower is better \| \| `latency` \| 1201.638 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 131072 provider Meta reasoning Not reported Reasoning effort max Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1 pp sourcePopulation 24 stderr 0.996 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_agent_benchmark.CFXFrRQl.json) Source Vals AI Original model label meta/muse-wonder Source version 1.1 Observed 2026-10-21T16:45:52.890Z Evidence SHA-256 `14681f4a2310e0f257df9ec07c9c81e5dadc8f6f71634f02e174a02a8096d868` Source locator `/tasks/criteria_pass_rate__white_collar_defense_investigations/meta~1muse-wonder` | | | | | | | | 2 | [Gemini 4 Argon](https://benchize.com/models/gemini-4-argon-high.md)High | 95.46% | SE ±1.21 pp | $10.94 | 1,880.17source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 95.455 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 10.940583 \| USD \| per_test \| Lower is better \| \| `latency` \| 1880.166 \| 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 Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.21 pp sourcePopulation 24 stderr 1.211 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_agent_benchmark.CFXFrRQl.json) Source Vals AI Original model label google/gemini-4-argon Source version 1.1 Observed 2026-10-21T16:45:52.890Z Evidence SHA-256 `14681f4a2310e0f257df9ec07c9c81e5dadc8f6f71634f02e174a02a8096d868` Source locator `/tasks/criteria_pass_rate__white_collar_defense_investigations/google~1gemini-4-argon` | | | | | | | | 3 | [Gemini 3.8 Flash](https://benchize.com/models/gemini-3-8-flash-high.md)High | 91.92% | SE ±1.58 pp | $3.12 | 1,076.18source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 91.919 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 3.121405 \| USD \| per_test \| Lower is better \| \| `latency` \| 1076.182 \| 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 Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.58 pp sourcePopulation 24 stderr 1.584 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_agent_benchmark.CFXFrRQl.json) Source Vals AI Original model label google/gemini-3.8-flash Source version 1.1 Observed 2026-10-21T16:45:52.890Z Evidence SHA-256 `14681f4a2310e0f257df9ec07c9c81e5dadc8f6f71634f02e174a02a8096d868` Source locator `/tasks/criteria_pass_rate__white_collar_defense_investigations/google~1gemini-3.8-flash` | | | | | | | | 4 | [MiniMax M3.1 Flash](https://benchize.com/models/minimax-m3-1-flash-max.md)Max | 91.41% | SE ±1.63 pp | $0.46 | 1,357.85source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 91.414 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.463122 \| USD \| per_test \| Lower is better \| \| `latency` \| 1357.85 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort max harness Not reported maxOutputTokens 262144 provider MiniMax reasoning Not reported Reasoning effort max Temperature 1 topP 0.95 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.63 pp sourcePopulation 24 stderr 1.628 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_agent_benchmark.CFXFrRQl.json) Source Vals AI Original model label minimax/MiniMax-M3.1-flash Source version 1.1 Observed 2026-10-21T16:45:52.890Z Evidence SHA-256 `14681f4a2310e0f257df9ec07c9c81e5dadc8f6f71634f02e174a02a8096d868` Source locator `/tasks/criteria_pass_rate__white_collar_defense_investigations/minimax~1MiniMax-M3.1-flash` | | | | | | | | 5 | [MiMo-V2.6-Flash](https://benchize.com/models/mimo-v2-6-flash.md) | 90.57% | SE ±1.7 pp | $0.07 | 917.02source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 90.572 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.07138 \| USD \| per_test \| Lower is better \| \| `latency` \| 917.02 \| 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: ±1.7 pp sourcePopulation 24 stderr 1.698 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_agent_benchmark.CFXFrRQl.json) Source Vals AI Original model label xiaomi/mimo-v2.6-flash Source version 1.1 Observed 2026-10-21T16:45:52.890Z Evidence SHA-256 `14681f4a2310e0f257df9ec07c9c81e5dadc8f6f71634f02e174a02a8096d868` Source locator `/tasks/criteria_pass_rate__white_collar_defense_investigations/xiaomi~1mimo-v2.6-flash` | | | | | | | | 6 | [MiMo-V2.6-Pro](https://benchize.com/models/mimo-v2-6-pro.md) | 90.07% | SE ±1.74 pp | $0.17 | 1,712.1source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 90.067 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.16563 \| USD \| per_test \| Lower is better \| \| `latency` \| 1712.104 \| 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: ±1.74 pp sourcePopulation 24 stderr 1.738 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_agent_benchmark.CFXFrRQl.json) Source Vals AI Original model label xiaomi/mimo-v2.6-pro Source version 1.1 Observed 2026-10-21T16:45:52.890Z Evidence SHA-256 `14681f4a2310e0f257df9ec07c9c81e5dadc8f6f71634f02e174a02a8096d868` Source locator `/tasks/criteria_pass_rate__white_collar_defense_investigations/xiaomi~1mimo-v2.6-pro` | | | | | | | | 7 | [DeepSeek V4.1 Flash 1017](https://benchize.com/models/deepseek-v4-1-flash-1017-max.md)Max | 89.9% | SE ±1.75 pp | $0.69 | 1,431.65source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 89.899 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.688859 \| USD \| per_test \| Lower is better \| \| `latency` \| 1431.654 \| 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 1 topP 1 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.75 pp sourcePopulation 24 stderr 1.752 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_agent_benchmark.CFXFrRQl.json) Source Vals AI Original model label deepseek/deepseek-v4.1-flash-1017 Source version 1.1 Observed 2026-10-21T16:45:52.890Z Evidence SHA-256 `14681f4a2310e0f257df9ec07c9c81e5dadc8f6f71634f02e174a02a8096d868` Source locator `/tasks/criteria_pass_rate__white_collar_defense_investigations/deepseek~1deepseek-v4.1-flash-1017` | | | | | | | | 8 | [Claude Sonnet 5.5](https://benchize.com/models/claude-sonnet-5-5-max.md)Max | 89.23% | SE ±1.8 pp | $17.21 | 4,937.71source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 89.226 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 17.206836 \| USD \| per_test \| Lower is better \| \| `latency` \| 4937.706 \| 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: ±1.8 pp sourcePopulation 24 stderr 1.802 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_agent_benchmark.CFXFrRQl.json) Source Vals AI Original model label anthropic/claude-sonnet-5-5 Source version 1.1 Observed 2026-10-21T16:45:52.890Z Evidence SHA-256 `14681f4a2310e0f257df9ec07c9c81e5dadc8f6f71634f02e174a02a8096d868` Source locator `/tasks/criteria_pass_rate__white_collar_defense_investigations/anthropic~1claude-sonnet-5-5` | | | | | | | | 9 | [GPT-6 Astra](https://benchize.com/models/gpt-6-astra-max.md)Max | 89.06% | SE ±1.81 pp | $17.86 | 1,732.53source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 89.057 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 17.857837 \| USD \| per_test \| Lower is better \| \| `latency` \| 1732.526 \| 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: ±1.81 pp sourcePopulation 24 stderr 1.814 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_agent_benchmark.CFXFrRQl.json) Source Vals AI Original model label openai/gpt-6-astra Source version 1.1 Observed 2026-10-21T16:45:52.890Z Evidence SHA-256 `14681f4a2310e0f257df9ec07c9c81e5dadc8f6f71634f02e174a02a8096d868` Source locator `/tasks/criteria_pass_rate__white_collar_defense_investigations/openai~1gpt-6-astra` | | | | | | | | 10 | [DeepSeek V4 Pro 0813](https://benchize.com/models/deepseek-v4-pro-0813-max.md)Max | 88.89% | SE ±1.83 pp | $0.62 | 477.32source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 88.889 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.622985 \| USD \| per_test \| Lower is better \| \| `latency` \| 477.316 \| 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: ±1.83 pp sourcePopulation 24 stderr 1.827 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_agent_benchmark.CFXFrRQl.json) Source Vals AI Original model label deepseek/deepseek-v4-pro-0813 Source version 1.1 Observed 2026-10-21T16:45:52.890Z Evidence SHA-256 `14681f4a2310e0f257df9ec07c9c81e5dadc8f6f71634f02e174a02a8096d868` Source locator `/tasks/criteria_pass_rate__white_collar_defense_investigations/deepseek~1deepseek-v4-pro-0813` | | | | | | | | 11 | [GPT-6.1 Sol](https://benchize.com/models/gpt-6-1-sol-max.md)Max | 87.21% | SE ±1.94 pp | $3.05 | 2,123.43source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 87.205 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 3.053435 \| USD \| per_test \| Lower is better \| \| `latency` \| 2123.426 \| 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: ±1.94 pp sourcePopulation 24 stderr 1.942 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_agent_benchmark.CFXFrRQl.json) Source Vals AI Original model label openai/gpt-6.1-sol Source version 1.1 Observed 2026-10-21T16:45:52.890Z Evidence SHA-256 `14681f4a2310e0f257df9ec07c9c81e5dadc8f6f71634f02e174a02a8096d868` Source locator `/tasks/criteria_pass_rate__white_collar_defense_investigations/openai~1gpt-6.1-sol` | | | | | | | | 12 | [GPT-6 Luna](https://benchize.com/models/gpt-6-luna-max.md)Max | 85.86% | SE ±2.03 pp | $0.27 | 1,480.51source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 85.859 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.268224 \| USD \| per_test \| Lower is better \| \| `latency` \| 1480.506 \| 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: ±2.03 pp sourcePopulation 24 stderr 2.025 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_agent_benchmark.CFXFrRQl.json) Source Vals AI Original model label openai/gpt-6-luna Source version 1.1 Observed 2026-10-21T16:45:52.890Z Evidence SHA-256 `14681f4a2310e0f257df9ec07c9c81e5dadc8f6f71634f02e174a02a8096d868` Source locator `/tasks/criteria_pass_rate__white_collar_defense_investigations/openai~1gpt-6-luna` | | | | | | | | 13 | [Claude Haiku 5.5](https://benchize.com/models/claude-haiku-5-5-max.md)Max | 84.34% | SE ±2.11 pp | $5.02 | 3,043.18source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 84.343 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 5.01937 \| USD \| per_test \| Lower is better \| \| `latency` \| 3043.182 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort max harness Not reported maxOutputTokens 128000 provider Anthropic reasoning Not reported Reasoning effort max Temperature Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±2.11 pp sourcePopulation 24 stderr 2.112 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_agent_benchmark.CFXFrRQl.json) Source Vals AI Original model label anthropic/claude-haiku-5-5 Source version 1.1 Observed 2026-10-21T16:45:52.890Z Evidence SHA-256 `14681f4a2310e0f257df9ec07c9c81e5dadc8f6f71634f02e174a02a8096d868` Source locator `/tasks/criteria_pass_rate__white_collar_defense_investigations/anthropic~1claude-haiku-5-5` | | | | | | | | 14 | [Claude Opus 5.5](https://benchize.com/models/claude-opus-5-5-max.md)Max | 83.33% | SE ±2.17 pp | $24.22 | 3,518.71source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 83.333 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 24.223778 \| USD \| per_test \| Lower is better \| \| `latency` \| 3518.708 \| 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: ±2.17 pp sourcePopulation 24 stderr 2.166 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_legal_agent_benchmark.CFXFrRQl.json) Source Vals AI Original model label anthropic/claude-opus-5-5 Source version 1.1 Observed 2026-10-21T16:45:52.890Z Evidence SHA-256 `14681f4a2310e0f257df9ec07c9c81e5dadc8f6f71634f02e174a02a8096d868` Source locator `/tasks/criteria_pass_rate__white_collar_defense_investigations/anthropic~1claude-opus-5-5` | | | | | | | --- Cite as: HLAB-Vals-1.1 / criteria pass rate white collar defense investigations — Benchize, https://benchize.com/benchmarks/vals-ai%3Ahlab%3Acriteria_pass_rate__white_collar_defense_investigations (retrieved 2026-10-21)