Consensus score
53.5
95% interval 47.6 to 57.7
53.5
95% interval 47.6 to 57.7
51th
IQR spread 26.4 pct
109
Across 8 domains
3
Newest observation 2d ago
Median within-board percentile by capability domain across boards with sufficient peer overlap.
Best verified run per board, with confidence intervals and Pareto frontier status.
| Benchmark | Rank | Score | Cost | Status | Observed |
|---|---|---|---|---|---|
| GDPval Epoch AI Benchmarking Hub · medium | #6 / 11 | 34.8% | Unknown | Standard | Not stated |
| LMArena Vision (style control) LMArena · Entity recognition · high | #6 / 52Tied #1 | 1,261[1,228 to 1,295] | Unknown | Standard | 9 Nov 2026 |
| METR time horizon METR time horizons · triframe_inspect/triframe_agent | #10 / 23Tied #1 | 203 min[112.6 min to 405.6 min] | Unknown | Standard | Not stated |
| HLE Epoch AI Benchmarking Hub · high | #15 / 56 | 25.3%[22% to 28.7%] | Unknown | Standard | Not stated |
| HLE Epoch AI Benchmarking Hub · Default | #15 / 56 | 25.3%[22% to 28.7%] | Unknown | Standard | Not stated |
| LMArena Vision (style control) LMArena · Captioning · high | #15 / 34 | 1,195[1,159 to 1,232] | Unknown | Standard | 9 Nov 2026 |
| LMArena Vision LMArena · Entity recognition · high | #16 / 52 | 1,261[1,231 to 1,292] | Unknown | Standard | 9 Nov 2026 |
| SWE-Bench verified Epoch AI Benchmarking Hub · high | #20 / 35 | 73.6%[69.6% to 77.5%] | Unknown | Standard | 6 Feb 2026 |
| LMArena Vision LMArena · Creative writing · Default | #21 / 34 | 1,189[1,173 to 1,204] | Unknown | Standard | 9 Jan 2026 |
| FrontierMath-2025-02-28-Private Epoch AI Benchmarking Hub · high | #22 / 101 | 32.4%[27% to 37.8%] | Unknown | Standard | 13 Nov 2025 |
| LMArena Vision LMArena · Captioning · high | #23 / 34 | 1,203[1,171 to 1,235] | Unknown | Standard | 9 Nov 2026 |
| LMArena Vision (style control) LMArena · Creative writing · Default | #24 / 34 | 1,176[1,158 to 1,194] | Unknown | Standard | 9 Jan 2026 |
| SWE-Bench verified Epoch AI Benchmarking Hub · medium | #24 / 35 | 71.5%[67.5% to 75.5%] | Unknown | Standard | 5 Feb 2026 |
| LMArena Search (style control) LMArena · Overall · search | #27 / 42 | 1,189[1,181 to 1,196] | Unknown | Standard | 9 Nov 2026 |
| FrontierMath-2025-02-28-Private Epoch AI Benchmarking Hub · medium | #28 / 101 | 27.2%[22.1% to 32.4%] | Unknown | Standard | 13 Nov 2025 |
| LMArena Search LMArena · Overall · search | #38 / 42 | 1,133[1,127 to 1,139] | Unknown | Standard | 9 Nov 2026 |
| LMArena Text (style control) LMArena · Japanese · high | #61 / 288 | 1,431[1,409 to 1,453] | Unknown | Standard | 11 Nov 2026 |
| LMArena Vision (style control) LMArena · Homework · high | #67 / 118 | 1,259[1,243 to 1,276] | Unknown | Standard | 9 Nov 2026 |
| Terminal Bench Epoch AI Benchmarking Hub · Codex CLI | #73 / 202 | 49.6%[43.9% to 55.3%] | Unknown | Standard | 7 Aug 2025 |
| LMArena Vision (style control) LMArena · Humor · high | #74 / 102 | 1,201[1,179 to 1,224] | Unknown | Standard | 9 Nov 2026 |
| Terminal Bench Epoch AI Benchmarking Hub · medium · Codex CLI | #74 / 202 | 49.6%[43.9% to 55.3%] | Unknown | Standard | 4 Nov 2025 |
| LMArena Vision (style control) LMArena · Diagram · high | #76 / 123 | 1,251[1,238 to 1,264] | Unknown | Standard | 9 Nov 2026 |
| LMArena Vision LMArena · Humor · high | #77 / 102 | 1,206[1,185 to 1,227] | Unknown | Standard | 9 Nov 2026 |
| LMArena Vision (style control) LMArena · OCR · high | #78 / 123 | 1,233[1,224 to 1,242] | Unknown | Standard | 9 Nov 2026 |
| LMArena Text LMArena · Japanese · high | #79 / 288 | 1,402[1,381 to 1,424] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Polish · high | #79 / 243 | 1,449[1,434 to 1,464] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Industry legal and government · high | #80 / 393 | 1,460[1,446 to 1,475] | Unknown | Standard | 11 Nov 2026 |
| Terminal Bench Epoch AI Benchmarking Hub · OpenHands | #83 / 202 | 43.8%[38% to 49.7%] | Unknown | Standard | 7 Aug 2025 |
| Terminal Bench Epoch AI Benchmarking Hub · medium · OpenHands | #84 / 202 | 43.8%[37.9% to 49.7%] | Unknown | Standard | 2 Nov 2025 |
| LMArena Vision LMArena · Chinese · high | #85 / 135 | 1,235[1,218 to 1,252] | Unknown | Standard | 9 Nov 2026 |
| LMArena Vision (style control) LMArena · Chinese · high | #85 / 135 | 1,229[1,211 to 1,247] | Unknown | Standard | 9 Nov 2026 |
| LMArena Text (style control) LMArena · German · high | #87 / 320 | 1,446[1,424 to 1,469] | Unknown | Standard | 11 Nov 2026 |
| LMArena Vision (style control) LMArena · English · high | #87 / 167 | 1,215[1,206 to 1,225] | Unknown | Standard | 9 Nov 2026 |
| LMArena Vision (style control) LMArena · Overall · high | #87 / 167 | 1,211[1,204 to 1,218] | Unknown | Standard | 9 Nov 2026 |
| LMArena Vision LMArena · Diagram · high | #88 / 123 | 1,226[1,214 to 1,238] | Unknown | Standard | 9 Nov 2026 |
| LMArena Vision LMArena · English · high | #90 / 167 | 1,218[1,209 to 1,227] | Unknown | Standard | 9 Nov 2026 |
| LMArena Vision LMArena · OCR · high | #90 / 123 | 1,215[1,206 to 1,224] | Unknown | Standard | 9 Nov 2026 |
| LMArena Vision LMArena · Homework · high | #91 / 118 | 1,227[1,211 to 1,243] | Unknown | Standard | 9 Nov 2026 |
| LMArena Vision LMArena · Overall · high | #91 / 167 | 1,208[1,201 to 1,215] | Unknown | Standard | 9 Nov 2026 |
| LMArena Vision LMArena · Creative writing vision · high | #92 / 106 | 1,184[1,165 to 1,202] | Unknown | Standard | 9 Nov 2026 |
| LMArena Text (style control) LMArena · Korean · high | #95 / 289 | 1,391[1,369 to 1,413] | Unknown | Standard | 11 Nov 2026 |
| LMArena Vision (style control) LMArena · Creative writing vision · high | #95 / 106 | 1,170[1,151 to 1,189] | Unknown | Standard | 9 Nov 2026 |
| LMArena Text (style control) LMArena · Industry medicine and healthcare · high | #96 / 391 | 1,460[1,445 to 1,475] | Unknown | Standard | 11 Nov 2026 |
| LMArena WebDev LMArena · Webdev HTML · medium | #96 / 155 | 1,426[1,409 to 1,443] | Unknown | Standard | 15 Nov 2026 |
| LMArena WebDev LMArena · Overall · medium | #98 / 156 | 1,417[1,402 to 1,433] | Unknown | Standard | 15 Nov 2026 |
| LMArena WebDev LMArena · Webdev · medium | #98 / 156 | 1,417[1,402 to 1,433] | Unknown | Standard | 15 Nov 2026 |
| GPQA diamond Epoch AI Benchmarking Hub · high | #99 / 338 | 86.2%[82.1% to 90.2%] | Unknown | Standard | 29 Oct 2025 |
| LMArena Text (style control) LMArena · Non English · high | #102 / 421 | 1,425[1,419 to 1,430] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Math · high | #103 / 404 | 1,435[1,421 to 1,449] | Unknown | Standard | 11 Nov 2026 |
| GPQA diamond Epoch AI Benchmarking Hub · medium | #106 / 338 | 85.4%[81.2% to 89.5%] | Unknown | Standard | 7 Aug 2025 |
| LMArena Text (style control) LMArena · Industry mathematical · high | #107 / 402 | 1,444[1,429 to 1,458] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Overall · high | #108 / 421 | 1,435[1,430 to 1,439] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Russian · high | #108 / 385 | 1,430[1,417 to 1,443] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · German · high | #110 / 320 | 1,416[1,394 to 1,438] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Exclude ties · high | #110 / 421 | 1,428[1,422 to 1,434] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Expert · high | #110 / 371 | 1,460[1,444 to 1,475] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Polish · high | #112 / 243 | 1,412[1,397 to 1,426] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Industry entertainment and sports and media · high | #115 / 419 | 1,398[1,390 to 1,407] | Unknown | Standard | 11 Nov 2026 |
| Terminal Bench Epoch AI Benchmarking Hub · medium · Terminus 2 | #117 / 202 | 35.2%[29.1% to 41.3%] | Unknown | Standard | 31 Oct 2025 |
| Terminal Bench Epoch AI Benchmarking Hub · Terminus 2 | #118 / 202 | 35.2%[29.2% to 41.2%] | Unknown | Standard | 7 Aug 2025 |
| LMArena Text (style control) LMArena · French · high | #119 / 301 | 1,444[1,415 to 1,473] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Hard prompts · high | #119 / 421 | 1,448[1,442 to 1,454] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · English · high | #120 / 421 | 1,439[1,433 to 1,445] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Industry life and physical and social science · high | #121 / 419 | 1,447[1,438 to 1,455] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Spanish · high | #124 / 304 | 1,417[1,397 to 1,438] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Hard prompts English · high | #125 / 419 | 1,454[1,446 to 1,461] | Unknown | Standard | 11 Nov 2026 |
| Terminal Bench Epoch AI Benchmarking Hub · Mini-SWE-Agent | #125 / 202 | 33.9%[28.2% to 39.6%] | Unknown | Standard | 7 Aug 2025 |
| LMArena Text LMArena · Industry legal and government · high | #126 / 393 | 1,424[1,410 to 1,438] | Unknown | Standard | 11 Nov 2026 |
| Terminal Bench Epoch AI Benchmarking Hub · medium · Mini-SWE-Agent | #126 / 202 | 33.9%[28.2% to 39.6%] | Unknown | Standard | 3 Nov 2025 |
| LMArena Text LMArena · Coding · high | #128 / 416 | 1,436[1,428 to 1,444] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Coding · high | #129 / 416 | 1,471[1,463 to 1,479] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Industry entertainment and sports and media · high | #131 / 419 | 1,379[1,371 to 1,388] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Korean · high | #132 / 289 | 1,358[1,337 to 1,380] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Russian · high | #133 / 385 | 1,402[1,390 to 1,415] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Multi turn · high | #134 / 419 | 1,421[1,412 to 1,430] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Instruction following · high | #136 / 421 | 1,410[1,403 to 1,417] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Chinese · high | #138 / 395 | 1,453[1,440 to 1,467] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Industry writing and literature and language · high | #138 / 420 | 1,396[1,388 to 1,404] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Expert · high | #140 / 371 | 1,419[1,405 to 1,434] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Industry software and IT services · high | #140 / 421 | 1,454[1,448 to 1,461] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Industry medicine and healthcare · high | #141 / 391 | 1,413[1,399 to 1,428] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Non English · high | #142 / 421 | 1,397[1,392 to 1,403] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Industry business and management and financial operations · high | #142 / 414 | 1,414[1,406 to 1,423] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Hard prompts · high | #143 / 421 | 1,416[1,410 to 1,422] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · French · high | #149 / 301 | 1,408[1,380 to 1,437] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Hard prompts English · high | #149 / 419 | 1,419[1,412 to 1,427] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Instruction following · high | #149 / 421 | 1,388[1,381 to 1,395] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Overall · high | #149 / 421 | 1,406[1,402 to 1,411] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Exclude ties · high | #151 / 421 | 1,387[1,381 to 1,393] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Creative writing · high | #152 / 419 | 1,365[1,355 to 1,374] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Industry writing and literature and language · high | #154 / 420 | 1,379[1,371 to 1,386] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Multi turn · high | #156 / 419 | 1,399[1,390 to 1,408] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Longer query · high | #156 / 399 | 1,413[1,405 to 1,421] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Industry software and IT services · high | #157 / 421 | 1,420[1,414 to 1,427] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text (style control) LMArena · Creative writing · high | #157 / 419 | 1,373[1,363 to 1,382] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · English · high | #160 / 421 | 1,409[1,403 to 1,415] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Math · high | #162 / 404 | 1,398[1,384 to 1,412] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Industry mathematical · high | #164 / 402 | 1,405[1,390 to 1,419] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Spanish · high | #164 / 304 | 1,388[1,368 to 1,408] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Chinese · high | #166 / 395 | 1,422[1,409 to 1,435] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Industry life and physical and social science · high | #169 / 419 | 1,403[1,395 to 1,412] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Longer query · high | #172 / 399 | 1,387[1,380 to 1,395] | Unknown | Standard | 11 Nov 2026 |
| LMArena Text LMArena · Industry business and management and financial operations · high | #175 / 414 | 1,378[1,369 to 1,386] | Unknown | Standard | 11 Nov 2026 |
| ARC-AGI-2 Epoch AI Benchmarking Hub · high | #183 / 299 | 9.9% | $0.73 | Standard | Not stated |
| ARC-AGI-2 Epoch AI Benchmarking Hub · medium | #189 / 299 | 7.5% | $0.45 | Standard | Not stated |
| GPQA diamond Epoch AI Benchmarking Hub · minimal | #194 / 338 | 71.7%[65.4% to 78%] | Unknown | Standard | 20 Jul 2026 |
| ARC-AGI-2 Epoch AI Benchmarking Hub · low | #241 / 299 | 1.9% | $0.19 | Standard | Not stated |
| ARC-AGI-2 Epoch AI Benchmarking Hub · low | #241 / 299 | 1.9% | $0.19 | Standard | Not stated |
| ARC-AGI-2 Epoch AI Benchmarking Hub · minimal | #281 / 299 | 0% | $0.06 | Standard | Not stated |
Same benchmark and harness measured across multiple reasoning effort settings.
Epoch AI Benchmarking Hub
0.0
$0.06
1.9
$0.19
7.5
$0.45
9.9
$0.73
Epoch AI Benchmarking Hub
27.2
32.4
Epoch AI Benchmarking Hub
71.7
85.4
86.2
Epoch AI Benchmarking Hub
71.5
73.6
Direct pairwise win rates for GPT-5 across shared benchmark boards.
15 shared boards
7% win
Head-to-head15 shared boards
7% win
Head-to-head14 shared boards
0% win
Head-to-head14 shared boards
0% win
Head-to-head13 shared boards
46% win
Head-to-head13 shared boards
62% win
Head-to-head13 shared boards
0% win
Head-to-head13 shared boards
31% win
Head-to-head13 shared boards
0% win
Head-to-head13 shared boards
62% win
Head-to-head