Acropolis

GPT-5 nano

OpenAIGrade B. Evidence grade B of A to D: good.ProprietaryReleased 7 Aug 2025Consistent across benchmarks
Compare with peers

Domain profile

Median within-board percentile by capability domain across boards with sufficient peer overlap.

  • Agentic (1 boards): 11 (scale 0 to 100)Agentic (1 boards)11
  • Coding (1 boards): 11 (scale 0 to 100)Coding (1 boards)11
  • General text (3 boards): 46 (scale 0 to 100)General text (3 boards)46
  • Math (1 boards): 43 (scale 0 to 100)Math (1 boards)43
  • Science (2 boards): 35 (scale 0 to 100)Science (2 boards)35
  • Vision (2 boards): 32 (scale 0 to 100)Vision (2 boards)32

Benchmark appearances

Best verified run per board, with confidence intervals and Pareto frontier status.

BenchmarkRankScoreCostStatusObserved
LMArena Vision

LMArena · Creative writing · Default

#33 / 34
1,117[1,077 to 1,158]UnknownStandard9 Jan 2026
LMArena Vision (style control)

LMArena · Creative writing · Default

#34 / 34
1,092[1,050 to 1,134]UnknownStandard9 Jan 2026
FrontierMath-2025-02-28-Private

Epoch AI Benchmarking Hub · high

#60 / 101
8.3%[5.1% to 11.5%]UnknownStandard30 Oct 2025
FrontierMath-2025-02-28-Private

Epoch AI Benchmarking Hub · medium

#62 / 101
7.2%[4.3% to 10.2%]UnknownStandard7 Aug 2025
LMArena Vision

LMArena · Humor · high

#100 / 102
1,109[1,063 to 1,156]UnknownStandard9 Nov 2026
LMArena Vision

LMArena · Chinese · high

#101 / 135
1,176[1,132 to 1,219]UnknownStandard9 Nov 2026
LMArena Vision (style control)

LMArena · Humor · high

#102 / 102
1,092[1,044 to 1,140]UnknownStandard9 Nov 2026
LMArena Vision (style control)

LMArena · Homework · high

#103 / 118
1,188[1,152 to 1,223]UnknownStandard9 Nov 2026
LMArena Vision (style control)

LMArena · Chinese · high

#108 / 135
1,152[1,109 to 1,196]UnknownStandard9 Nov 2026
LMArena Vision

LMArena · Homework · high

#112 / 118
1,160[1,125 to 1,195]UnknownStandard9 Nov 2026
LMArena Vision (style control)

LMArena · OCR · high

#112 / 123
1,159[1,144 to 1,175]UnknownStandard9 Nov 2026
LMArena Vision

LMArena · English · high

#114 / 167
1,170[1,154 to 1,185]UnknownStandard9 Nov 2026
LMArena Vision

LMArena · OCR · high

#114 / 123
1,152[1,137 to 1,167]UnknownStandard9 Nov 2026
LMArena Vision

LMArena · Diagram · high

#115 / 123
1,139[1,113 to 1,165]UnknownStandard9 Nov 2026
LMArena Vision

LMArena · Overall · high

#115 / 167
1,159[1,148 to 1,170]UnknownStandard9 Nov 2026
LMArena Vision (style control)

LMArena · Diagram · high

#115 / 123
1,154[1,128 to 1,181]UnknownStandard9 Nov 2026
LMArena Vision (style control)

LMArena · Overall · high

#120 / 167
1,147[1,136 to 1,159]UnknownStandard9 Nov 2026
LMArena Vision (style control)

LMArena · English · high

#121 / 167
1,154[1,138 to 1,169]UnknownStandard9 Nov 2026
Terminal Bench

Epoch AI Benchmarking Hub · spoox-o-m

#169 / 202
21.8%[16.3% to 27.3%]UnknownStandard7 Aug 2025
LMArena Text (style control)

LMArena · Spanish · high

#172 / 304
1,383[1,344 to 1,423]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Spanish · high

#182 / 304
1,362[1,322 to 1,401]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · German · high

#188 / 320
1,348[1,307 to 1,390]UnknownStandard11 Nov 2026
Terminal Bench

Epoch AI Benchmarking Hub · medium · Codex CLI

#190 / 202
11.5%[7% to 16%]UnknownStandard4 Nov 2025
LMArena Text

LMArena · German · high

#191 / 320
1,326[1,285 to 1,367]UnknownStandard11 Nov 2026
Terminal Bench

Epoch AI Benchmarking Hub · Codex CLI

#191 / 202
11.5%[7% to 15.9%]UnknownStandard7 Aug 2025
Terminal Bench

Epoch AI Benchmarking Hub · medium · OpenHands

#192 / 202
9.9%[5.8% to 14%]UnknownStandard2 Nov 2025
Terminal Bench

Epoch AI Benchmarking Hub · OpenHands

#193 / 202
9.9%[5.7% to 14%]UnknownStandard7 Aug 2025
LMArena Text

LMArena · Korean · high

#194 / 289
1,269[1,229 to 1,310]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Polish · high

#194 / 243
1,313[1,288 to 1,338]UnknownStandard11 Nov 2026
Terminal Bench

Epoch AI Benchmarking Hub · medium · Terminus 2

#195 / 202
7.9%[4.2% to 11.6%]UnknownStandard31 Oct 2025
LMArena Text (style control)

LMArena · Polish · high

#196 / 243
1,338[1,312 to 1,363]UnknownStandard11 Nov 2026
Terminal Bench

Epoch AI Benchmarking Hub · Terminus 2

#196 / 202
7.9%[4.2% to 11.5%]UnknownStandard7 Aug 2025
Terminal Bench

Epoch AI Benchmarking Hub · medium · Mini-SWE-Agent

#197 / 202
7%[3.3% to 10.7%]UnknownStandard3 Nov 2025
Terminal Bench

Epoch AI Benchmarking Hub · Mini-SWE-Agent

#198 / 202
7%[3.2% to 10.7%]UnknownStandard7 Aug 2025
LMArena Text (style control)

LMArena · Korean · high

#202 / 289
1,279[1,238 to 1,321]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Japanese · high

#203 / 288
1,227[1,177 to 1,276]UnknownStandard11 Nov 2026
GPQA diamond

Epoch AI Benchmarking Hub · high

#204 / 338
69.4%[64% to 74.9%]UnknownStandard30 Oct 2025
GPQA diamond

Epoch AI Benchmarking Hub · medium

#209 / 338
67.4%[62.2% to 72.6%]UnknownStandard7 Aug 2025
LMArena Text (style control)

LMArena · Japanese · high

#210 / 288
1,245[1,195 to 1,295]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Chinese · high

#215 / 395
1,379[1,352 to 1,405]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Chinese · high

#216 / 395
1,356[1,330 to 1,382]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry mathematical · high

#219 / 402
1,356[1,327 to 1,384]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Math · high

#219 / 404
1,348[1,321 to 1,374]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Expert · high

#223 / 371
1,355[1,323 to 1,388]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry medicine and healthcare · high

#224 / 391
1,366[1,338 to 1,393]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Non English · high

#224 / 421
1,328[1,319 to 1,337]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Expert · high

#225 / 371
1,321[1,289 to 1,353]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry medicine and healthcare · high

#226 / 391
1,334[1,307 to 1,361]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Math · high

#226 / 404
1,318[1,292 to 1,345]UnknownStandard11 Nov 2026
ARC-AGI-2

Epoch AI Benchmarking Hub · high

#229 / 299
2.6%$0.03On frontier. Evidence grade A of A to D: strongest.Not stated
LMArena Text

LMArena · Coding · high

#230 / 416
1,351[1,337 to 1,366]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry mathematical · high

#230 / 402
1,326[1,298 to 1,355]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Non English · high

#230 / 421
1,313[1,304 to 1,322]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Hard prompts · high

#232 / 421
1,328[1,319 to 1,338]UnknownStandard11 Nov 2026
GPQA diamond

Epoch AI Benchmarking Hub · low

#233 / 338
57.6%[50.7% to 64.5%]UnknownStandard13 Jul 2026
LMArena Text

LMArena · Instruction following · high

#233 / 421
1,306[1,293 to 1,319]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Hard prompts English · high

#235 / 419
1,338[1,324 to 1,353]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry software and IT services · high

#235 / 421
1,345[1,334 to 1,356]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Coding · high

#235 / 416
1,385[1,371 to 1,400]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Hard prompts English · high

#235 / 419
1,370[1,356 to 1,385]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Instruction following · high

#235 / 421
1,326[1,313 to 1,339]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry business and management and financial operations · high

#236 / 414
1,311[1,295 to 1,326]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Hard prompts · high

#236 / 421
1,356[1,346 to 1,366]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Overall · high

#237 / 421
1,320[1,313 to 1,327]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry business and management and financial operations · high

#237 / 414
1,340[1,324 to 1,355]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Exclude ties · high

#238 / 421
1,268[1,259 to 1,278]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Russian · high

#240 / 385
1,296[1,273 to 1,320]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Exclude ties · high

#240 / 421
1,293[1,283 to 1,303]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry software and IT services · high

#240 / 421
1,372[1,361 to 1,383]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Longer query · high

#241 / 399
1,312[1,297 to 1,326]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Overall · high

#241 / 421
1,338[1,331 to 1,344]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Multi turn · high

#243 / 419
1,311[1,295 to 1,326]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry life and physical and social science · high

#244 / 419
1,318[1,302 to 1,335]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry life and physical and social science · high

#248 / 419
1,348[1,332 to 1,365]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Multi turn · high

#248 / 419
1,324[1,309 to 1,340]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry entertainment and sports and media · high

#250 / 419
1,268[1,252 to 1,284]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Russian · high

#251 / 385
1,312[1,288 to 1,335]UnknownStandard11 Nov 2026
LMArena Text

LMArena · English · high

#252 / 421
1,321[1,312 to 1,331]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry writing and literature and language · high

#254 / 420
1,283[1,269 to 1,297]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Longer query · high

#255 / 399
1,332[1,317 to 1,346]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · English · high

#257 / 421
1,344[1,334 to 1,354]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry legal and government · high

#259 / 393
1,297[1,270 to 1,324]UnknownStandard11 Nov 2026
GPQA diamond

Epoch AI Benchmarking Hub · minimal

#262 / 338
48.5%[41.5% to 55.5%]UnknownStandard13 Jul 2026
ARC-AGI-2

Epoch AI Benchmarking Hub · medium

#267 / 299
0.9%$0.01StandardNot stated
LMArena Text (style control)

LMArena · Industry legal and government · high

#269 / 393
1,327[1,300 to 1,355]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry entertainment and sports and media · high

#271 / 419
1,277[1,260 to 1,293]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Creative writing · high

#272 / 419
1,249[1,230 to 1,269]UnknownStandard11 Nov 2026
ARC-AGI-2

Epoch AI Benchmarking Hub · low

#281 / 299
0%$0.00StandardNot stated
ARC-AGI-2

Epoch AI Benchmarking Hub · minimal

#281 / 299
0%$0.00On frontier. Evidence grade A of A to D: strongest.Not stated
LMArena Text (style control)

LMArena · Industry writing and literature and language · high

#282 / 420
1,288[1,274 to 1,302]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Creative writing · high

#303 / 419
1,246[1,226 to 1,266]UnknownStandard11 Nov 2026

Reasoning effort scaling

Same benchmark and harness measured across multiple reasoning effort settings.

  • ARC-AGI-2

    Epoch AI Benchmarking Hub

    Monotone. Evidence grade A of A to D: strongest.
    minimal

    0.0

    $0.00

    low

    0.0

    $0.00

    medium

    0.9

    $0.01

    high

    2.6

    $0.03

  • FrontierMath-2025-02-28-Private

    Epoch AI Benchmarking Hub

    Monotone. Evidence grade A of A to D: strongest.
    medium

    7.2

    high

    8.3

  • GPQA diamond

    Epoch AI Benchmarking Hub

    Monotone. Evidence grade A of A to D: strongest.
    minimal

    48.5

    low

    57.6

    medium

    67.4

    high

    69.4

Closest head-to-head rivals

Direct pairwise win rates for GPT-5 nano across shared benchmark boards.