# MedQA / overall > Source: Benchize (https://benchize.com) — https://benchize.com/benchmarks/vals-ai%3Amedqa%3Aoverall. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/benchmarks/vals-ai%3Amedqa%3Aoverall. Evaluating language model bias in medical questions. Canonical URL: https://benchize.com/benchmarks/vals-ai%3Amedqa%3Aoverall 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 model bias in medical questions. Professional work · Models tested: 12 · Score publishers: Vals AI ## Model results | # | Model | Score | ± | Cost | Latency | | | --- | --- | --- | --- | --- | --- | --- | | 1 | [Gemini 3.1 Pro Preview](https://benchize.com/models/gemini-3-1-pro-preview-high.md)High | 96.37% | SE ±0.17 pp | $0.02 | 38.67source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 96.367 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.017689 \| USD \| per_test \| Lower is better \| \| `latency` \| 38.67 \| 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: ±0.17 pp sourcePopulation 95 stderr 0.171 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_medqa.Cco1kvog.json) Source Vals AI Original model label google/gemini-3.1-pro-preview Source version 1 Observed 2026-10-21T16:46:00.691Z Evidence SHA-256 `d67905978d932f64698ebe9e230d9519e5bae1c9d7176dda8379dc3874b6f90f` Source locator `/tasks/overall/google~1gemini-3.1-pro-preview` | | | | | | | | 2 | [Gemini 3 Flash Preview](https://benchize.com/models/gemini-3-flash-preview-high.md)High | 95.81% | SE ±0.18 pp | $0.01 | 26.32source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 95.808 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.012699 \| USD \| per_test \| Lower is better \| \| `latency` \| 26.324 \| 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: ±0.18 pp sourcePopulation 95 stderr 0.183 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_medqa.Cco1kvog.json) Source Vals AI Original model label google/gemini-3-flash-preview Source version 1 Observed 2026-10-21T16:46:00.691Z Evidence SHA-256 `d67905978d932f64698ebe9e230d9519e5bae1c9d7176dda8379dc3874b6f90f` Source locator `/tasks/overall/google~1gemini-3-flash-preview` | | | | | | | | 3 | [MiniMax M2.5](https://benchize.com/models/minimax-m2-5.md) | 92.53% | SE ±0.24 pp | $0.00293 | 33.54source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 92.525 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.002928 \| USD \| per_test \| Lower is better \| \| `latency` \| 33.54 \| 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: ±0.24 pp sourcePopulation 95 stderr 0.24 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_medqa.Cco1kvog.json) Source Vals AI Original model label minimax/MiniMax-M2.5 Source version 1 Observed 2026-10-21T16:46:00.691Z Evidence SHA-256 `d67905978d932f64698ebe9e230d9519e5bae1c9d7176dda8379dc3874b6f90f` Source locator `/tasks/overall/minimax~1MiniMax-M2.5` | | | | | | | | 4 | [Claude Sonnet 4.6](https://benchize.com/models/claude-sonnet-4-6-thinking.md)Thinking | 92.06% | SE ±0.25 pp | $0.04 | 35.12source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 92.058 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.035415 \| USD \| per_test \| Lower is better \| \| `latency` \| 35.119 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 128000 provider Anthropic reasoning true Reasoning effort unknown Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.25 pp sourcePopulation 95 stderr 0.247 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_medqa.Cco1kvog.json) Source Vals AI Original model label anthropic/claude-sonnet-4-6 Source version 1 Observed 2026-10-21T16:46:00.691Z Evidence SHA-256 `d67905978d932f64698ebe9e230d9519e5bae1c9d7176dda8379dc3874b6f90f` Source locator `/tasks/overall/anthropic~1claude-sonnet-4-6` | | | | | | | | 5 | [MiniMax M2.1](https://benchize.com/models/minimax-m2-1.md) | 91.16% | SE ±0.26 pp | $0.00238 | 35.94source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 91.158 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.002383 \| USD \| per_test \| Lower is better \| \| `latency` \| 35.938 \| 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: ±0.26 pp sourcePopulation 95 stderr 0.259 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_medqa.Cco1kvog.json) Source Vals AI Original model label minimax/MiniMax-M2.1 Source version 1 Observed 2026-10-21T16:46:00.691Z Evidence SHA-256 `d67905978d932f64698ebe9e230d9519e5bae1c9d7176dda8379dc3874b6f90f` Source locator `/tasks/overall/minimax~1MiniMax-M2.1` | | | | | | | | 6 | [GPT-4o (2024-08-06)](https://benchize.com/models/gpt-4o-2024-08-06-high.md)High | 88.16% | SE ±0.58 pp | $0.00279 | 3.39source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 88.161 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.002787 \| USD \| per_test \| Lower is better \| \| `latency` \| 3.392 \| 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: ±0.58 pp sourcePopulation 95 stderr 0.578 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_medqa.Cco1kvog.json) Source Vals AI Original model label openai/gpt-4o-2024-08-06 Source version 1 Observed 2026-10-21T16:46:00.691Z Evidence SHA-256 `d67905978d932f64698ebe9e230d9519e5bae1c9d7176dda8379dc3874b6f90f` Source locator `/tasks/overall/openai~1gpt-4o-2024-08-06` | | | | | | | | 7 | [Mistral Large 3 2512](https://benchize.com/models/mistral-large-2512.md) | 82.23% | SE ±0.35 pp | $0.000771 | 17.31source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 82.233 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.000771 \| USD \| per_test \| Lower is better \| \| `latency` \| 17.31 \| 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: ±0.35 pp sourcePopulation 95 stderr 0.349 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_medqa.Cco1kvog.json) Source Vals AI Original model label mistralai/mistral-large-2512 Source version 1 Observed 2026-10-21T16:46:00.691Z Evidence SHA-256 `d67905978d932f64698ebe9e230d9519e5bae1c9d7176dda8379dc3874b6f90f` Source locator `/tasks/overall/mistralai~1mistral-large-2512` | | | | | | | | 8 | [GPT-4 Turbo](https://benchize.com/models/gpt-4-turbo-high.md)High | 81.99% | SE ±0.69 pp | $0.01 | 8.41source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 81.986 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.012152 \| USD \| per_test \| Lower is better \| \| `latency` \| 8.413 \| 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: ±0.69 pp sourcePopulation 95 stderr 0.686 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_medqa.Cco1kvog.json) Source Vals AI Original model label openai/gpt-4-turbo Source version 1 Observed 2026-10-21T16:46:00.691Z Evidence SHA-256 `d67905978d932f64698ebe9e230d9519e5bae1c9d7176dda8379dc3874b6f90f` Source locator `/tasks/overall/openai~1gpt-4-turbo` | | | | | | | | 9 | [Command A](https://benchize.com/models/command-a.md) | 80.55% | SE ±0.36 pp | $0.00346 | 2.91source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 80.55 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.003459 \| USD \| per_test \| Lower is better \| \| `latency` \| 2.908 \| 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: ±0.36 pp sourcePopulation 95 stderr 0.361 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_medqa.Cco1kvog.json) Source Vals AI Original model label cohere/command-a-03-2025 Source version 1 Observed 2026-10-21T16:46:00.691Z Evidence SHA-256 `d67905978d932f64698ebe9e230d9519e5bae1c9d7176dda8379dc3874b6f90f` Source locator `/tasks/overall/cohere~1command-a-03-2025` | | | | | | | | 10 | [Gemini 2.5 Flash Lite Preview 09-2025](https://benchize.com/models/gemini-2-5-flash-lite-preview-09-2025.md) | 80.33% | SE ±0.36 pp | $0.000281 | 5.15source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 80.325 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.000281 \| USD \| per_test \| Lower is better \| \| `latency` \| 5.145 \| 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: ±0.36 pp sourcePopulation 95 stderr 0.363 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_medqa.Cco1kvog.json) Source Vals AI Original model label google/gemini-2.5-flash-lite-preview-09-2025 Source version 1 Observed 2026-10-21T16:46:00.691Z Evidence SHA-256 `d67905978d932f64698ebe9e230d9519e5bae1c9d7176dda8379dc3874b6f90f` Source locator `/tasks/overall/google~1gemini-2.5-flash-lite-preview-09-2025` | | | | | | | | 11 | [GPT-4o-mini (2024-07-18)](https://benchize.com/models/gpt-4o-mini-2024-07-18-high.md)High | 72.44% | SE ±0.8 pp | $0.000157 | 2.32source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 72.436 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.000157 \| USD \| per_test \| Lower is better \| \| `latency` \| 2.322 \| 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: ±0.8 pp sourcePopulation 95 stderr 0.799 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_medqa.Cco1kvog.json) Source Vals AI Original model label openai/gpt-4o-mini-2024-07-18 Source version 1 Observed 2026-10-21T16:46:00.691Z Evidence SHA-256 `d67905978d932f64698ebe9e230d9519e5bae1c9d7176dda8379dc3874b6f90f` Source locator `/tasks/overall/openai~1gpt-4o-mini-2024-07-18` | | | | | | | | 12 | [GPT-3.5 Turbo](https://benchize.com/models/gpt-3-5-turbo-high.md)High | 58.47% | SE ±0.88 pp | $0.000424 | 1.59source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 58.471 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.000424 \| USD \| per_test \| Lower is better \| \| `latency` \| 1.589 \| 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: ±0.88 pp sourcePopulation 95 stderr 0.881 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_medqa.Cco1kvog.json) Source Vals AI Original model label openai/gpt-3.5-turbo Source version 1 Observed 2026-10-21T16:46:00.691Z Evidence SHA-256 `d67905978d932f64698ebe9e230d9519e5bae1c9d7176dda8379dc3874b6f90f` Source locator `/tasks/overall/openai~1gpt-3.5-turbo` | | | | | | | --- Cite as: MedQA / overall — Benchize, https://benchize.com/benchmarks/vals-ai%3Amedqa%3Aoverall (retrieved 2026-10-21)