MMMU-Pro
AA · Vision understanding · Objective
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #156 · Source label: GPT-4o (Aug '24)
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Artificial Analysis
- Raw value
- 56.3%
- Percentile
- 25.8%
- Last updated
- recent
- Eligibility
- headline eligible
Parsed from Artificial Analysis public leaderboard field `mmmuPro`.
25.8% percentile inside its fair comparison set56.3%Raw benchmark value
Vision Arena
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #54 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,239
- Percentile
- 61.9%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: overall. Source rank: #61. Votes: 23416. Organization: openai. License: Proprietary.
61.9% percentile inside its fair comparison set1,239Raw benchmark valueCI 1,233 - 1,244
Vision Arena · Captioning
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #10 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,212
- Percentile
- 71%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: captioning. Source rank: #10. Votes: 273. Organization: openai. License: Proprietary.
71% percentile inside its fair comparison set1,212Raw benchmark valueCI 1,176 - 1,248
Vision Arena · Creative Writing Vision
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #36 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,249
- Percentile
- 57.3%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: creative_writing_vision. Source rank: #41. Votes: 1245. Organization: openai. License: Proprietary.
57.3% percentile inside its fair comparison set1,249Raw benchmark valueCI 1,231 - 1,267
Vision Arena · Diagram
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #56 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,261
- Percentile
- 44.4%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: diagram. Source rank: #62. Votes: 3749. Organization: openai. License: Proprietary.
44.4% percentile inside its fair comparison set1,261Raw benchmark valueCI 1,251 - 1,272
Vision Arena · English
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #57 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,236
- Percentile
- 59.4%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: english. Source rank: #63. Votes: 10183. Organization: openai. License: Proprietary.
59.4% percentile inside its fair comparison set1,236Raw benchmark valueCI 1,229 - 1,244
Vision Arena · Entity Recognition
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #19 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,218
- Percentile
- 55%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: entity_recognition. Source rank: #21. Votes: 295. Organization: openai. License: Proprietary.
55% percentile inside its fair comparison set1,218Raw benchmark valueCI 1,186 - 1,250
Vision Arena · Homework
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #65 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,245
- Percentile
- 33.3%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: homework. Source rank: #74. Votes: 2123. Organization: openai. License: Proprietary.
33.3% percentile inside its fair comparison set1,245Raw benchmark valueCI 1,232 - 1,258
Vision Arena · Humor
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #42 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,225
- Percentile
- 43.8%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: humor. Source rank: #47. Votes: 795. Organization: openai. License: Proprietary.
43.8% percentile inside its fair comparison set1,225Raw benchmark valueCI 1,203 - 1,247
Vision Arena · Ocr
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #56 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,247
- Percentile
- 45%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: ocr. Source rank: #63. Votes: 11180. Organization: openai. License: Proprietary.
45% percentile inside its fair comparison set1,247Raw benchmark valueCI 1,240 - 1,254
Vision Arena · No Style Control
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #61 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,243
- Percentile
- 56.8%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: overall. Source rank: #67. Votes: 23416. Organization: openai. License: Proprietary.
56.8% percentile inside its fair comparison set1,243Raw benchmark valueCI 1,238 - 1,249
Vision Arena · Captioning · No Style Control
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #15 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,223
- Percentile
- 54.8%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: captioning. Source rank: #16. Votes: 273. Organization: openai. License: Proprietary.
54.8% percentile inside its fair comparison set1,223Raw benchmark valueCI 1,188 - 1,259
Vision Arena · Creative Writing Vision · No Style Control
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #42 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,254
- Percentile
- 50%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: creative_writing_vision. Source rank: #47. Votes: 1245. Organization: openai. License: Proprietary.
50% percentile inside its fair comparison set1,254Raw benchmark valueCI 1,236 - 1,271
Vision Arena · Diagram · No Style Control
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #58 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,255
- Percentile
- 42.4%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: diagram. Source rank: #64. Votes: 3749. Organization: openai. License: Proprietary.
42.4% percentile inside its fair comparison set1,255Raw benchmark valueCI 1,245 - 1,266
Vision Arena · English · No Style Control
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #60 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,244
- Percentile
- 57.2%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: english. Source rank: #65. Votes: 10183. Organization: openai. License: Proprietary.
57.2% percentile inside its fair comparison set1,244Raw benchmark valueCI 1,237 - 1,252
Vision Arena · Entity Recognition · No Style Control
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #19 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,245
- Percentile
- 55%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: entity_recognition. Source rank: #21. Votes: 295. Organization: openai. License: Proprietary.
55% percentile inside its fair comparison set1,245Raw benchmark valueCI 1,213 - 1,276
Vision Arena · Homework · No Style Control
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #62 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,252
- Percentile
- 36.5%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: homework. Source rank: #69. Votes: 2123. Organization: openai. License: Proprietary.
36.5% percentile inside its fair comparison set1,252Raw benchmark valueCI 1,239 - 1,265
Vision Arena · Humor · No Style Control
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #39 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,248
- Percentile
- 47.9%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: humor. Source rank: #44. Votes: 795. Organization: openai. License: Proprietary.
47.9% percentile inside its fair comparison set1,248Raw benchmark valueCI 1,226 - 1,270
Vision Arena · Ocr · No Style Control
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #60 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,249
- Percentile
- 41%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: ocr. Source rank: #66. Votes: 11180. Organization: openai. License: Proprietary.
41% percentile inside its fair comparison set1,249Raw benchmark valueCI 1,242 - 1,256
MMMU Pro
VALS-AI · Vision understanding · Objective
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #72 · Source label: openai/gpt-4o-2024-08-06
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Vals AI
- Raw value
- 64%
- Percentile
- 14.5%
- Last updated
- aging
- Eligibility
- headline eligible
Parsed from Vals AI BenchmarkView overall scores. Vals slug: mmmu; provider: OpenAI.
14.5% percentile inside its fair comparison set64%Raw benchmark valueCI 61.7% - 66.3%
Vision Arena · Creative Writing
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #8 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,249
- Percentile
- 78.1%
- Last updated
- archived
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: creative_writing. Source rank: #9. Votes: 1047. Organization: openai. License: Proprietary.
78.1% percentile inside its fair comparison set1,249Raw benchmark valueCI 1,231 - 1,268
Vision Arena · Creative Writing · No Style Control
AR · Vision understanding · Human
It is useful when the model must read charts, UI, screenshots, or visual scenes rather than text alone.
Rank #9 · Source label: chatgpt-4o-latest-20250326
verified runtimeexact alias
Raw row drilldownsource row, percentile, last updated, eligibility
- Source
- Arena
- Raw value
- 1,248
- Percentile
- 75%
- Last updated
- archived
- Eligibility
- headline eligible
Parsed from Arena leaderboard dataset row `chatgpt-4o-latest-20250326`. Category: creative_writing. Source rank: #11. Votes: 1047. Organization: openai. License: Proprietary.
75% percentile inside its fair comparison set1,248Raw benchmark valueCI 1,230 - 1,266