# MMMU-Pro > Source: Benchize (https://benchize.com) — https://benchize.com/benchmarks/vals-ai%3Ammmu%3Aoverall. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/benchmarks/vals-ai%3Ammmu%3Aoverall. Multimodal understanding on MMMU-Pro without Python augmentation. Canonical URL: https://benchize.com/benchmarks/vals-ai%3Ammmu%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. Multimodal understanding on MMMU-Pro without Python augmentation. Multimodal · Models tested: 29 · Score publishers: [moonshot k3](https://www.kimi.ai/blog/kimi-k3), Vals AI ## Model results | # | Model | Score | ± | Cost | Latency | Source | | | --- | --- | --- | --- | --- | --- | --- | --- | | 1 / 28 | [Claude Fable 5.1](https://benchize.com/models/claude-fable-5-1-max.md)Max | 90.64% | SE ±0.7 pp | $0.17 | 37.47source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 90.636 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.166285 \| USD \| per_test \| Lower is better \| \| `latency` \| 37.47 \| 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: ±0.7 pp sourcePopulation 93 stderr 0.7 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label anthropic/claude-fable-5-1 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/anthropic~1claude-fable-5-1` | | | | | | | | | 2 / 28 | [Claude Opus 5](https://benchize.com/models/claude-opus-5-max.md)Max | 89.88% | SE ±0.73 pp | $0.07 | 30.03source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 89.884 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.069014 \| USD \| per_test \| Lower is better \| \| `latency` \| 30.028 \| 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: ±0.73 pp sourcePopulation 93 stderr 0.725 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label anthropic/claude-opus-5 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/anthropic~1claude-opus-5` | | | | | | | | | 3 / 28 | [Claude Fable 5](https://benchize.com/models/claude-fable-5-max.md)Max | 89.31% | SE ±0.74 pp | $0.16 | 61.44source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 89.306 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.16064 \| USD \| per_test \| Lower is better \| \| `latency` \| 61.44 \| 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: ±0.74 pp sourcePopulation 93 stderr 0.743 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label anthropic/claude-fable-5 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/anthropic~1claude-fable-5` | | | | | | | | | 4 / 28 | [Gemini 3.8 Flash](https://benchize.com/models/gemini-3-8-flash-high.md)High | 89.08% | SE ±0.75 pp | $0.00447 | 12.03source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 89.075 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.004472 \| USD \| per_test \| Lower is better \| \| `latency` \| 12.028 \| 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: ±0.75 pp sourcePopulation 93 stderr 0.75 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label google/gemini-3.8-flash Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/google~1gemini-3.8-flash` | | | | | | | | | 5 / 28 | [Gemini 3.7 Flash](https://benchize.com/models/gemini-3-7-flash-high.md)High | 88.96% | SE ±0.75 pp | $0.01 | 6.28source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 88.96 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.014052 \| USD \| per_test \| Lower is better \| \| `latency` \| 6.283 \| 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: ±0.75 pp sourcePopulation 93 stderr 0.753 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label google/gemini-3.7-flash Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/google~1gemini-3.7-flash` | | | | | | | | | 6 / 28 | [GPT-5.6 Sol](https://benchize.com/models/gpt-5-6-sol-max.md)Max | 88.84% | SE ±0.76 pp | $0.06 | 26.2source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 88.844 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.055795 \| USD \| per_test \| Lower is better \| \| `latency` \| 26.198 \| 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: ±0.76 pp sourcePopulation 93 stderr 0.757 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label openai/gpt-5.6-sol Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/openai~1gpt-5.6-sol` | | | | | | | | | 7 / 28 | [Gemini 3.6 Flash](https://benchize.com/models/gemini-3-6-flash-high.md)High | 88.38% | SE ±0.77 pp | $0.02 | 11.31source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 88.382 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.0186 \| USD \| per_test \| Lower is better \| \| `latency` \| 11.313 \| 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: ±0.77 pp sourcePopulation 93 stderr 0.77 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label google/gemini-3.6-flash Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/google~1gemini-3.6-flash` | | | | | | | | | 8 / 28 | [GPT-5.5](https://benchize.com/models/gpt-5-5-xhigh.md)Extra high | 88.27% | SE ±0.77 pp | $0.04 | 54.15source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 88.266 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.043319 \| USD \| per_test \| Lower is better \| \| `latency` \| 54.147 \| 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: ±0.77 pp sourcePopulation 93 stderr 0.774 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label openai/gpt-5.5 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/openai~1gpt-5.5` | | | | | | | | | 8 / 28 | [Gemini 3.5 Flash](https://benchize.com/models/gemini-3-5-flash-high.md)High | 88.27% | SE ±0.77 pp | $0.03 | 12.23source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 88.266 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.025155 \| USD \| per_test \| Lower is better \| \| `latency` \| 12.226 \| 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: ±0.77 pp sourcePopulation 93 stderr 0.774 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label google/gemini-3.5-flash Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/google~1gemini-3.5-flash` | | | | | | | | | 10 / 28 | [Gemini 3.1 Pro Preview](https://benchize.com/models/gemini-3-1-pro-preview-high.md)High | 88.21% | SE ±0.78 pp | $0.06 | 76.99source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 88.208 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.064084 \| USD \| per_test \| Lower is better \| \| `latency` \| 76.99 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 65535 provider Google reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0.78 pp sourcePopulation 93 stderr 0.775 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label google/gemini-3.1-pro-preview Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/google~1gemini-3.1-pro-preview` | | | | | | | | | 11 / 28 | [Gemini 3 Flash Preview](https://benchize.com/models/gemini-3-flash-preview-high.md)High | 87.63% | SE ±0.79 pp | $0.01 | 27.86source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 87.63 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.014422 \| USD \| per_test \| Lower is better \| \| `latency` \| 27.856 \| 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: ±0.79 pp sourcePopulation 93 stderr 0.792 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label google/gemini-3-flash-preview Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/google~1gemini-3-flash-preview` | | | | | | | | | 12 / 28 | [Claude Opus 4.8](https://benchize.com/models/claude-opus-4-8-max.md)Max | 86.59% | SE ±0.82 pp | $0.12 | 49.04source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 86.59 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.120663 \| USD \| per_test \| Lower is better \| \| `latency` \| 49.044 \| 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: ±0.82 pp sourcePopulation 93 stderr 0.819 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label anthropic/claude-opus-4-8 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/anthropic~1claude-opus-4-8` | | | | | | | | | 12 / 28 | [Muse Spark 1.1](https://benchize.com/models/muse-spark-1-1-xhigh.md)Extra high | 86.59% | SE ±0.82 pp | $0.01 | 31.72source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 86.59 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.01361 \| USD \| per_test \| Lower is better \| \| `latency` \| 31.715 \| 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: ±0.82 pp sourcePopulation 93 stderr 0.819 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label meta/muse_spark_1_1 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/meta~1muse_spark_1_1` | | | | | | | | | 14 / 28 | [GPT-5.6 Terra](https://benchize.com/models/gpt-5-6-terra-max.md)Max | 86.47% | SE ±0.82 pp | $0.02 | 13.03source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 86.474 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.016375 \| USD \| per_test \| Lower is better \| \| `latency` \| 13.033 \| 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: ±0.82 pp sourcePopulation 93 stderr 0.822 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label openai/gpt-5.6-terra Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/openai~1gpt-5.6-terra` | | | | | | | | | 15 / 28 | [Muse Spark 1.2](https://benchize.com/models/muse-spark-1-2-xhigh.md)Extra high | 86.13% | SE ±0.83 pp | $0.02 | 73.84source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 86.127 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.018022 \| USD \| per_test \| Lower is better \| \| `latency` \| 73.839 \| 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: ±0.83 pp sourcePopulation 93 stderr 0.831 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label meta/muse_spark_1_2 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/meta~1muse_spark_1_2` | | | | | | | | | 16 / 28 | [Claude Opus 4.7](https://benchize.com/models/claude-opus-4-7-max.md)Max | 85.55% | SE ±0.85 pp | $0.17 | 70.8source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 85.549 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.170461 \| USD \| per_test \| Lower is better \| \| `latency` \| 70.801 \| 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: ±0.85 pp sourcePopulation 93 stderr 0.845 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label anthropic/claude-opus-4-7 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/anthropic~1claude-opus-4-7` | | | | | | | | | 17 / 28 | [GPT-5.6 Luna](https://benchize.com/models/gpt-5-6-luna-max.md)Max | 85.03% | SE ±0.86 pp | $0.00642 | 35.1source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 85.029 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.006419 \| USD \| per_test \| Lower is better \| \| `latency` \| 35.102 \| 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: ±0.86 pp sourcePopulation 93 stderr 0.858 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label openai/gpt-5.6-luna Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/openai~1gpt-5.6-luna` | | | | | | | | | 18 / 28 | [Claude Sonnet 4.6](https://benchize.com/models/claude-sonnet-4-6-max.md)Max | 83.58% | SE ±0.89 pp | $0.14 | 131.73source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 83.584 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.144122 \| USD \| per_test \| Lower is better \| \| `latency` \| 131.728 \| 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: ±0.89 pp sourcePopulation 93 stderr 0.891 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label anthropic/claude-sonnet-4-6 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/anthropic~1claude-sonnet-4-6` | | | | | | | | | 18 / 28 | [Gemini 3.5 Flash-Lite](https://benchize.com/models/gemini-3-5-flash-lite-high.md)High | 83.58% | SE ±0.89 pp | $0.00658 | 7.8source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 83.584 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.006577 \| USD \| per_test \| Lower is better \| \| `latency` \| 7.799 \| 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: ±0.89 pp sourcePopulation 93 stderr 0.891 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label google/gemini-3.5-flash-lite Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/google~1gemini-3.5-flash-lite` | | | | | | | | | 20 / 28 | [Claude Sonnet 5](https://benchize.com/models/claude-sonnet-5-max.md)Max | 83.01% | SE ±0.9 pp | $0.03 | 18.82source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 83.006 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.033133 \| USD \| per_test \| Lower is better \| \| `latency` \| 18.817 \| 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: ±0.9 pp sourcePopulation 93 stderr 0.903 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label anthropic/claude-sonnet-5 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/anthropic~1claude-sonnet-5` | | | | | | | | | 1 / 5 | [GPT-5.6 Sol](https://benchize.com/models/gpt-5-6-sol-max.md)Max | 83% | — | — | — | [Moonshot AI · sha256:0e806bbad76dc8e8dbc899a86051e3406bdd5995fe6c6aa1b2176dde4fb28e6c](https://www.kimi.ai/blog/kimi-k3) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 83.0 \| percent \| reported \| Higher is better \| ### Configuration imageOrder original_images_prepended Reasoning effort max runs 3 thinkingEnabled true toolsEnabled false ### Reported statistics runCount 3 ### Source Score publisher [Moonshot AI](https://www.kimi.ai/blog/kimi-k3) Collected via [moonshot k3](https://www.kimi.ai/blog/kimi-k3) Publisher reference Referenced by the collection source; the upstream result has not been independently captured. Benchmark [https://www.kimi.ai/blog/kimi-k3](https://www.kimi.ai/blog/kimi-k3) [Collected via](https://www.kimi.ai/blog/kimi-k3) Source Moonshot AI Original model label GPT 5.6 Sol (max) Source version sha256:0e806bbad76dc8e8dbc899a86051e3406bdd5995fe6c6aa1b2176dde4fb28e6c Observed 2026-10-21T16:31:44.130Z Evidence SHA-256 `0e806bbad76dc8e8dbc899a86051e3406bdd5995fe6c6aa1b2176dde4fb28e6c` Source locator `/flight/sheet/data/23/scores/2` | | | | | | | | | 21 / 28 | [Gemini 3.1 Flash Lite Preview](https://benchize.com/models/gemini-3-1-flash-lite-preview-high.md)High | 82.49% | SE ±0.91 pp | $0.00319 | 12.45source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 82.486 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.003187 \| USD \| per_test \| Lower is better \| \| `latency` \| 12.448 \| 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: ±0.91 pp sourcePopulation 93 stderr 0.914 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label google/gemini-3.1-flash-lite-preview Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/google~1gemini-3.1-flash-lite-preview` | | | | | | | | | 2 / 5 | [Kimi K3](https://benchize.com/models/kimi-k3-max.md)Max | 81.6% | — | — | — | [Moonshot AI · sha256:0e806bbad76dc8e8dbc899a86051e3406bdd5995fe6c6aa1b2176dde4fb28e6c](https://www.kimi.ai/blog/kimi-k3) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 81.6 \| percent \| reported \| Higher is better \| ### Configuration imageOrder original_images_prepended Reasoning effort max runs 3 Temperature 1 thinkingEnabled true toolsEnabled false topP 1 ### Reported statistics runCount 3 ### Source Score publisher [Moonshot AI](https://www.kimi.ai/blog/kimi-k3) Collected via [moonshot k3](https://www.kimi.ai/blog/kimi-k3) Publisher reference Referenced by the collection source; the upstream result has not been independently captured. Benchmark [https://www.kimi.ai/blog/kimi-k3](https://www.kimi.ai/blog/kimi-k3) [Collected via](https://www.kimi.ai/blog/kimi-k3) Source Moonshot AI Original model label Kimi K3 (max) Source version sha256:0e806bbad76dc8e8dbc899a86051e3406bdd5995fe6c6aa1b2176dde4fb28e6c Observed 2026-10-21T16:31:44.130Z Evidence SHA-256 `0e806bbad76dc8e8dbc899a86051e3406bdd5995fe6c6aa1b2176dde4fb28e6c` Source locator `/flight/sheet/data/23/scores/0` | | | | | | | | | 3 / 5 | [GPT-5.5](https://benchize.com/models/gpt-5-5-xhigh.md)Extra high | 81.2% | — | — | — | [Moonshot AI · sha256:0e806bbad76dc8e8dbc899a86051e3406bdd5995fe6c6aa1b2176dde4fb28e6c](https://www.kimi.ai/blog/kimi-k3) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 81.2 \| percent \| reported \| Higher is better \| ### Configuration imageOrder original_images_prepended Reasoning effort xhigh runs 3 thinkingEnabled true toolsEnabled false ### Reported statistics runCount 3 ### Source Score publisher [Moonshot AI](https://www.kimi.ai/blog/kimi-k3) Collected via [moonshot k3](https://www.kimi.ai/blog/kimi-k3) Publisher reference Referenced by the collection source; the upstream result has not been independently captured. Benchmark [https://www.kimi.ai/blog/kimi-k3](https://www.kimi.ai/blog/kimi-k3) [Collected via](https://www.kimi.ai/blog/kimi-k3) Source Moonshot AI Original model label GPT 5.5 (xhigh) Source version sha256:0e806bbad76dc8e8dbc899a86051e3406bdd5995fe6c6aa1b2176dde4fb28e6c Observed 2026-10-21T16:31:44.130Z Evidence SHA-256 `0e806bbad76dc8e8dbc899a86051e3406bdd5995fe6c6aa1b2176dde4fb28e6c` Source locator `/flight/sheet/data/23/scores/4` | | | | | | | | | 3 / 5 | [Claude Fable 5](https://benchize.com/models/claude-fable-5-max.md)Max | 81.2% | — | — | — | [Moonshot AI · sha256:0e806bbad76dc8e8dbc899a86051e3406bdd5995fe6c6aa1b2176dde4fb28e6c](https://www.kimi.ai/blog/kimi-k3) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 81.2 \| percent \| reported \| Higher is better \| ### Configuration Fallback policy unspecified fallbackEnabled true imageOrder original_images_prepended Reasoning effort max runs 3 sourceFallbackLabel with fallback thinkingEnabled true toolsEnabled false Fallback is enabled for refusals; this does not mean every task used a fallback model. ### Reported statistics runCount 3 ### Source Score publisher [Moonshot AI](https://www.kimi.ai/blog/kimi-k3) Collected via [moonshot k3](https://www.kimi.ai/blog/kimi-k3) Publisher reference Referenced by the collection source; the upstream result has not been independently captured. Benchmark [https://www.kimi.ai/blog/kimi-k3](https://www.kimi.ai/blog/kimi-k3) [Collected via](https://www.kimi.ai/blog/kimi-k3) Source Moonshot AI Original model label Claude Fable 5 (max, with fallback) Source version sha256:0e806bbad76dc8e8dbc899a86051e3406bdd5995fe6c6aa1b2176dde4fb28e6c Observed 2026-10-21T16:31:44.130Z Evidence SHA-256 `0e806bbad76dc8e8dbc899a86051e3406bdd5995fe6c6aa1b2176dde4fb28e6c` Source locator `/flight/sheet/data/23/scores/1` | | | | | | | | | 22 / 28 | [MiniMax M3](https://benchize.com/models/minimax-m3.md) | 81.16% | SE ±0.94 pp | $0.00943 | 68.35source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 81.156 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.009431 \| USD \| per_test \| Lower is better \| \| `latency` \| 68.354 \| 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: ±0.94 pp sourcePopulation 93 stderr 0.94 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label minimax/MiniMax-M3 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/minimax~1MiniMax-M3` | | | | | | | | | 23 / 28 | [MiMo-V2.5](https://benchize.com/models/mimo-v2-5.md) | 80% | SE ±0.96 pp | $0.0014 | 42.47source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 80 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.001396 \| USD \| per_test \| Lower is better \| \| `latency` \| 42.47 \| 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: ±0.96 pp sourcePopulation 93 stderr 0.962 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label xiaomi/mimo-v2.5 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/xiaomi~1mimo-v2.5` | | | | | | | | | 5 / 5 | [Claude Opus 4.8](https://benchize.com/models/claude-opus-4-8-max.md)Max | 78.9% | — | — | — | [Moonshot AI · sha256:0e806bbad76dc8e8dbc899a86051e3406bdd5995fe6c6aa1b2176dde4fb28e6c](https://www.kimi.ai/blog/kimi-k3) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 78.9 \| percent \| reported \| Higher is better \| ### Configuration imageOrder original_images_prepended Reasoning effort max runs 3 thinkingEnabled true toolsEnabled false ### Reported statistics runCount 3 ### Source Score publisher [Moonshot AI](https://www.kimi.ai/blog/kimi-k3) Collected via [moonshot k3](https://www.kimi.ai/blog/kimi-k3) Publisher reference Referenced by the collection source; the upstream result has not been independently captured. Benchmark [https://www.kimi.ai/blog/kimi-k3](https://www.kimi.ai/blog/kimi-k3) [Collected via](https://www.kimi.ai/blog/kimi-k3) Source Moonshot AI Original model label Claude Opus 4.8 (max) Source version sha256:0e806bbad76dc8e8dbc899a86051e3406bdd5995fe6c6aa1b2176dde4fb28e6c Observed 2026-10-21T16:31:44.130Z Evidence SHA-256 `0e806bbad76dc8e8dbc899a86051e3406bdd5995fe6c6aa1b2176dde4fb28e6c` Source locator `/flight/sheet/data/23/scores/3` | | | | | | | | | 24 / 28 | [Gemini 2.5 Flash Lite Preview 09-2025](https://benchize.com/models/gemini-2-5-flash-lite-preview-09-2025.md) | 72.54% | SE ±1.07 pp | $0.000899 | 20.07source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 72.543 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.000899 \| USD \| per_test \| Lower is better \| \| `latency` \| 20.069 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 65536 provider Google reasoning Not reported Reasoning effort unknown Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.07 pp sourcePopulation 93 stderr 1.073 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label google/gemini-2.5-flash-lite-preview-09-2025 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/google~1gemini-2.5-flash-lite-preview-09-2025` | | | | | | | | | 25 / 28 | [Mistral Large 3 2512](https://benchize.com/models/mistral-large-2512.md) | 66.19% | SE ±1.14 pp | $0.00148 | 13.55source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 66.185 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.001477 \| USD \| per_test \| Lower is better \| \| `latency` \| 13.55 \| 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: ±1.14 pp sourcePopulation 93 stderr 1.137 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label mistralai/mistral-large-2512 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/mistralai~1mistral-large-2512` | | | | | | | | | 26 / 28 | [GPT-4o (2024-08-06)](https://benchize.com/models/gpt-4o-2024-08-06-high.md)High | 64.01% | SE ±1.15 pp | $0.00625 | 47.52source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 64.009 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.006251 \| USD \| per_test \| Lower is better \| \| `latency` \| 47.52 \| 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: ±1.15 pp sourcePopulation 93 stderr 1.154 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label openai/gpt-4o-2024-08-06 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/openai~1gpt-4o-2024-08-06` | | | | | | | | | 27 / 28 | [GPT-4o (2024-11-20)](https://benchize.com/models/gpt-4o-2024-11-20-high.md)High | 62.16% | SE ±1.17 pp | $0.00682 | 45.87source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 62.161 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.006822 \| USD \| per_test \| Lower is better \| \| `latency` \| 45.873 \| 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: ±1.17 pp sourcePopulation 93 stderr 1.166 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label openai/gpt-4o-2024-11-20 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/openai~1gpt-4o-2024-11-20` | | | | | | | | | 28 / 28 | [GPT-4o-mini (2024-07-18)](https://benchize.com/models/gpt-4o-mini-2024-07-18-high.md)High | 56.56% | SE ±1.19 pp | $0.00037 | 45.02source duration | [Vals AI · 1](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 56.557 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.00037 \| USD \| per_test \| Lower is better \| \| `latency` \| 45.016 \| 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: ±1.19 pp sourcePopulation 93 stderr 1.191 tokenTotals Not reported ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_mmmu.BBln8x5x.json) Source Vals AI Original model label openai/gpt-4o-mini-2024-07-18 Source version 1 Observed 2026-10-21T16:46:07.000Z Evidence SHA-256 `29395186eddd2c70601228c584a14b3c1d8f9dba4b68256561b46b0dd03c635a` Source locator `/tasks/overall/openai~1gpt-4o-mini-2024-07-18` | | | | | | | | --- Cite as: MMMU-Pro — Benchize, https://benchize.com/benchmarks/vals-ai%3Ammmu%3Aoverall (retrieved 2026-10-21)