Acropolis

Gemini 3.5 Flash

GoogleGrade B. Evidence grade B of A to D: good.ProprietaryReleased 19 May 2026Within-domain variance
Compare with peers

Domain profile

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

  • Agentic (2 boards): 74 (scale 0 to 100)Agentic (2 boards)74
  • Coding (2 boards): 74 (scale 0 to 100)Coding (2 boards)74
  • Finance (1 boards): 55 (scale 0 to 100)Finance (1 boards)55
  • General text (5 boards): 74 (scale 0 to 100)General text (5 boards)74
  • Math (1 boards): 86 (scale 0 to 100)Math (1 boards)86
  • Science (2 boards): 81 (scale 0 to 100)Science (2 boards)81
  • Vision (2 boards): 91 (scale 0 to 100)Vision (2 boards)91

Benchmark appearances

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

BenchmarkRankScoreCostStatusObserved
SWE-Bench verified

Epoch AI Benchmarking Hub · high

#3 / 35Tied #1
79.3%[75.7% to 82.9%]UnknownStandard1 Jun 2026
LMArena Text

LMArena · Polish · high

#6 / 243Tied #1
1,515[1,494 to 1,537]UnknownStandard11 Nov 2026
LMArena Vision

LMArena · Humor · medium

#7 / 102Tied #1
1,323[1,298 to 1,348]UnknownStandard9 Nov 2026
LMArena Vision

LMArena · Chinese · medium

#8 / 135Tied #1
1,376[1,355 to 1,397]UnknownStandard9 Nov 2026
LMArena Vision

LMArena · Humor · high

#8 / 102Tied #1
1,321[1,297 to 1,346]UnknownStandard9 Nov 2026
LMArena Vision (style control)

LMArena · Humor · medium

#8 / 102
1,296[1,271 to 1,321]UnknownStandard9 Nov 2026
LMArena Text (style control)

LMArena · Polish · high

#9 / 243Tied #1
1,505[1,483 to 1,526]UnknownStandard11 Nov 2026
LMArena Vision

LMArena · Chinese · high

#9 / 135
1,373[1,352 to 1,394]UnknownStandard9 Nov 2026
LMArena Vision

LMArena · English · high

#13 / 167
1,305[1,295 to 1,314]UnknownStandard9 Nov 2026
LMArena Vision

LMArena · OCR · medium

#13 / 123
1,315[1,308 to 1,322]UnknownStandard9 Nov 2026
LMArena Vision

LMArena · Overall · high

#13 / 167
1,308[1,301 to 1,314]UnknownStandard9 Nov 2026
LMArena Vision (style control)

LMArena · Humor · high

#13 / 102
1,293[1,269 to 1,317]UnknownStandard9 Nov 2026
FrontierMath-2025-02-28-Private

Epoch AI Benchmarking Hub · high

#14 / 101
39%[33.3% to 44.6%]UnknownStandard22 May 2026
LMArena Text

LMArena · Math · high

#14 / 404
1,504[1,492 to 1,517]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Polish · medium

#14 / 243Tied #1
1,498[1,476 to 1,521]UnknownStandard11 Nov 2026
LMArena Vision

LMArena · Diagram · medium

#14 / 123
1,322[1,311 to 1,332]UnknownStandard9 Nov 2026
LMArena Vision (style control)

LMArena · Chinese · medium

#14 / 135Tied #1
1,345[1,323 to 1,366]UnknownStandard9 Nov 2026
LMArena Text

LMArena · Russian · high

#15 / 385
1,494[1,485 to 1,502]UnknownStandard11 Nov 2026
LMArena Vision

LMArena · OCR · high

#15 / 123
1,311[1,304 to 1,318]UnknownStandard9 Nov 2026
LMArena Vision

LMArena · Overall · medium

#15 / 167
1,305[1,299 to 1,312]UnknownStandard9 Nov 2026
LMArena Text

LMArena · German · high

#16 / 320
1,491[1,469 to 1,514]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Japanese · high

#16 / 288
1,489[1,465 to 1,513]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Korean · medium

#16 / 289
1,455[1,434 to 1,477]UnknownStandard11 Nov 2026
LMArena Vision

LMArena · Diagram · high

#16 / 123
1,321[1,311 to 1,331]UnknownStandard9 Nov 2026
LMArena Vision

LMArena · Homework · medium

#16 / 118
1,328[1,313 to 1,342]UnknownStandard9 Nov 2026
LMArena Vision (style control)

LMArena · Chinese · high

#16 / 135Tied #1
1,341[1,320 to 1,363]UnknownStandard9 Nov 2026
LMArena Text

LMArena · German · medium

#17 / 320
1,491[1,468 to 1,513]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Non English · high

#17 / 421
1,477[1,472 to 1,482]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Russian · medium

#17 / 385
1,493[1,484 to 1,501]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry writing and literature and language · high

#18 / 420
1,479[1,472 to 1,485]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Spanish · high

#18 / 304
1,482[1,465 to 1,498]UnknownStandard11 Nov 2026
LMArena Vision

LMArena · English · medium

#18 / 167
1,298[1,288 to 1,307]UnknownStandard9 Nov 2026
LMArena Text

LMArena · Creative writing · medium

#19 / 419
1,471[1,463 to 1,478]UnknownStandard11 Nov 2026
LMArena Vision

LMArena · Creative writing vision · high

#19 / 106
1,299[1,282 to 1,317]UnknownStandard9 Nov 2026
LMArena Text

LMArena · Creative writing · high

#20 / 419
1,469[1,462 to 1,476]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Exclude ties · high

#20 / 421
1,492[1,487 to 1,497]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Overall · high

#20 / 421
1,483[1,479 to 1,487]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Spanish · medium

#20 / 304
1,477[1,460 to 1,495]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Math · high

#20 / 404
1,497[1,485 to 1,510]UnknownStandard11 Nov 2026
LMArena Vision

LMArena · Creative writing vision · medium

#20 / 106
1,299[1,281 to 1,317]UnknownStandard9 Nov 2026
LMArena Vision (style control)

LMArena · Homework · medium

#20 / 118
1,314[1,300 to 1,329]UnknownStandard9 Nov 2026
LMArena Text

LMArena · Industry writing and literature and language · medium

#21 / 420
1,473[1,466 to 1,479]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Japanese · high

#21 / 288
1,482[1,458 to 1,506]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Polish · medium

#21 / 243
1,496[1,474 to 1,519]UnknownStandard11 Nov 2026
LMArena Vision

LMArena · Homework · high

#21 / 118
1,323[1,309 to 1,337]UnknownStandard9 Nov 2026
LMArena Document

LMArena · Overall · high

#22 / 46
1,462[1,453 to 1,471]UnknownStandard9 Nov 2026
LMArena Text

LMArena · Industry entertainment and sports and media · high

#22 / 419
1,459[1,453 to 1,466]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry mathematical · medium

#22 / 402
1,495[1,482 to 1,507]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Korean · high

#22 / 289
1,449[1,429 to 1,469]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Creative writing · medium

#22 / 419
1,469[1,462 to 1,477]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry writing and literature and language · high

#22 / 420
1,475[1,469 to 1,481]UnknownStandard11 Nov 2026
LMArena Vision (style control)

LMArena · OCR · medium

#22 / 123
1,298[1,291 to 1,305]UnknownStandard9 Nov 2026
LMArena Document

LMArena · Overall · medium

#23 / 46
1,461[1,453 to 1,468]UnknownStandard9 Nov 2026
LMArena Text

LMArena · English · high

#23 / 421
1,480[1,475 to 1,485]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Hard prompts · high

#23 / 421
1,489[1,485 to 1,494]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry entertainment and sports and media · medium

#23 / 419
1,459[1,452 to 1,466]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Longer query · high

#23 / 399
1,484[1,478 to 1,489]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry legal and government · high

#24 / 393
1,487[1,477 to 1,496]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry life and physical and social science · medium

#24 / 419
1,495[1,488 to 1,503]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Math · medium

#24 / 404
1,487[1,474 to 1,500]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Non English · medium

#24 / 421
1,472[1,467 to 1,477]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Russian · medium

#24 / 385
1,492[1,483 to 1,501]UnknownStandard11 Nov 2026
LMArena Vision (style control)

LMArena · Creative writing vision · medium

#24 / 106
1,275[1,257 to 1,292]UnknownStandard9 Nov 2026
LMArena Vision (style control)

LMArena · Overall · medium

#24 / 167
1,281[1,275 to 1,288]UnknownStandard9 Nov 2026
LMArena Text

LMArena · Exclude ties · medium

#25 / 421
1,486[1,481 to 1,491]UnknownStandard11 Nov 2026
LMArena Text

LMArena · French · high

#25 / 301
1,491[1,474 to 1,507]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry life and physical and social science · high

#25 / 419
1,493[1,486 to 1,501]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry medicine and healthcare · high

#25 / 391
1,484[1,473 to 1,495]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Multi turn · high

#25 / 419
1,480[1,473 to 1,488]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Overall · medium

#25 / 421
1,478[1,474 to 1,482]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Creative writing · high

#25 / 419
1,465[1,458 to 1,473]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · German · medium

#25 / 320
1,489[1,466 to 1,511]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Russian · high

#25 / 385
1,491[1,483 to 1,500]UnknownStandard11 Nov 2026
LMArena Vision (style control)

LMArena · Creative writing vision · high

#25 / 106
1,272[1,254 to 1,290]UnknownStandard9 Nov 2026
LMArena Vision (style control)

LMArena · Diagram · medium

#25 / 123
1,311[1,301 to 1,321]UnknownStandard9 Nov 2026
LMArena Vision (style control)

LMArena · Overall · high

#25 / 167
1,281[1,274 to 1,288]UnknownStandard9 Nov 2026
LMArena Text

LMArena · Chinese · medium

#26 / 395
1,527[1,516 to 1,538]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry medicine and healthcare · medium

#26 / 391
1,483[1,472 to 1,494]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry entertainment and sports and media · medium

#26 / 419
1,457[1,450 to 1,464]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry writing and literature and language · medium

#26 / 420
1,472[1,465 to 1,478]UnknownStandard11 Nov 2026
LMArena Vision (style control)

LMArena · English · high

#26 / 167
1,277[1,268 to 1,287]UnknownStandard9 Nov 2026
LMArena Text

LMArena · Industry business and management and financial operations · high

#27 / 414
1,467[1,460 to 1,474]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry legal and government · medium

#27 / 393
1,485[1,475 to 1,496]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Korean · medium

#27 / 289
1,446[1,425 to 1,468]UnknownStandard11 Nov 2026
LMArena Document (style control)

LMArena · Overall · medium

#28 / 46
1,459[1,451 to 1,467]UnknownStandard9 Nov 2026
LMArena Text

LMArena · Industry software and IT services · high

#28 / 421
1,494[1,489 to 1,499]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Japanese · medium

#28 / 288
1,461[1,437 to 1,486]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Non English · high

#28 / 421
1,470[1,465 to 1,475]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Spanish · high

#28 / 304
1,474[1,457 to 1,491]UnknownStandard11 Nov 2026
LMArena Document (style control)

LMArena · Overall · high

#29 / 46
1,455[1,446 to 1,464]UnknownStandard9 Nov 2026
LMArena Text

LMArena · Chinese · high

#29 / 395
1,526[1,515 to 1,536]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · German · high

#29 / 320
1,484[1,462 to 1,506]UnknownStandard11 Nov 2026
LMArena Vision (style control)

LMArena · OCR · high

#29 / 123
1,291[1,284 to 1,298]UnknownStandard9 Nov 2026
LMArena Text

LMArena · Hard prompts · medium

#30 / 421
1,485[1,480 to 1,490]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry mathematical · high

#30 / 402
1,491[1,479 to 1,502]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Multi turn · medium

#30 / 419
1,477[1,469 to 1,484]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry entertainment and sports and media · high

#30 / 419
1,455[1,448 to 1,461]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry life and physical and social science · medium

#30 / 419
1,497[1,489 to 1,504]UnknownStandard11 Nov 2026
LMArena Vision (style control)

LMArena · Diagram · high

#31 / 123
1,307[1,297 to 1,318]UnknownStandard9 Nov 2026
Vending-Bench 2

Epoch AI Benchmarking Hub · Default

#31 / 72
$5,396.42UnknownStandardNot stated
GPQA diamond

Epoch AI Benchmarking Hub · high

#32 / 338Tied #1
92.8%[89.6% to 96%]UnknownStandard22 May 2026
LMArena Text (style control)

LMArena · Exclude ties · high

#32 / 421
1,488[1,483 to 1,493]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Overall · high

#32 / 421
1,478[1,474 to 1,482]UnknownStandard11 Nov 2026
LMArena Vision (style control)

LMArena · Homework · high

#32 / 118
1,307[1,293 to 1,321]UnknownStandard9 Nov 2026
LMArena Text

LMArena · English · medium

#33 / 421
1,476[1,470 to 1,481]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Exclude ties · medium

#33 / 421
1,487[1,481 to 1,492]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Hard prompts English · high

#34 / 419
1,483[1,476 to 1,489]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · French · high

#34 / 301
1,494[1,478 to 1,511]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Non English · medium

#34 / 421
1,468[1,463 to 1,473]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Overall · medium

#34 / 421
1,476[1,472 to 1,480]UnknownStandard11 Nov 2026
LMArena Vision (style control)

LMArena · English · medium

#34 / 167
1,274[1,264 to 1,283]UnknownStandard9 Nov 2026
LMArena Text

LMArena · Expert · high

#35 / 371
1,494[1,485 to 1,503]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Korean · high

#35 / 289
1,440[1,419 to 1,460]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Spanish · medium

#35 / 304
1,471[1,454 to 1,488]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Coding · high

#36 / 416
1,493[1,487 to 1,499]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Instruction following · high

#36 / 421
1,467[1,462 to 1,473]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry medicine and healthcare · medium

#36 / 391
1,488[1,477 to 1,499]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Longer query · medium

#37 / 399
1,477[1,472 to 1,482]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Math · medium

#37 / 404
1,482[1,469 to 1,495]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry business and management and financial operations · medium

#38 / 414
1,459[1,452 to 1,466]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Chinese · medium

#38 / 395
1,518[1,507 to 1,529]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry legal and government · medium

#38 / 393
1,482[1,471 to 1,492]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry mathematical · medium

#38 / 402
1,489[1,477 to 1,501]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Japanese · medium

#39 / 288
1,459[1,434 to 1,484]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Instruction following · medium

#40 / 421
1,465[1,459 to 1,471]UnknownStandard11 Nov 2026
LMArena Text

LMArena · French · medium

#42 / 301
1,473[1,456 to 1,490]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Industry software and IT services · medium

#42 / 421
1,486[1,480 to 1,491]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Longer query · high

#42 / 399
1,483[1,478 to 1,489]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Multi turn · high

#42 / 419
1,480[1,472 to 1,487]UnknownStandard11 Nov 2026
LMArena WebDev

LMArena · Image to webdev · medium

#42 / 64
1,505[1,493 to 1,517]UnknownStandard15 Nov 2026
LMArena Text

LMArena · Coding · medium

#43 / 416
1,487[1,481 to 1,494]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Hard prompts English · medium

#43 / 419
1,479[1,472 to 1,485]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Hard prompts · high

#43 / 421
1,496[1,491 to 1,501]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry life and physical and social science · high

#43 / 419
1,489[1,482 to 1,497]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Chinese · high

#45 / 395
1,514[1,503 to 1,525]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Hard prompts · medium

#45 / 421
1,495[1,491 to 1,500]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry legal and government · high

#45 / 393
1,479[1,468 to 1,489]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Instruction following · medium

#45 / 421
1,466[1,460 to 1,472]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Multi turn · medium

#45 / 419
1,478[1,471 to 1,486]UnknownStandard11 Nov 2026
LMArena WebDev

LMArena · Image to webdev · high

#45 / 64
1,500[1,486 to 1,514]UnknownStandard15 Nov 2026
LMArena Text (style control)

LMArena · English · high

#47 / 421
1,478[1,473 to 1,483]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry business and management and financial operations · high

#47 / 414
1,468[1,461 to 1,475]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry medicine and healthcare · high

#47 / 391
1,486[1,475 to 1,497]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Instruction following · high

#47 / 421
1,465[1,460 to 1,471]UnknownStandard11 Nov 2026
LMArena Text

LMArena · Expert · medium

#48 / 371
1,485[1,477 to 1,494]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry mathematical · high

#48 / 402
1,483[1,471 to 1,495]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Longer query · medium

#49 / 399
1,480[1,475 to 1,486]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · English · medium

#50 / 421
1,477[1,472 to 1,482]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Hard prompts English · medium

#52 / 419
1,490[1,484 to 1,497]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Hard prompts English · high

#55 / 419
1,490[1,484 to 1,496]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry software and IT services · high

#57 / 421
1,502[1,496 to 1,507]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry business and management and financial operations · medium

#58 / 414
1,463[1,456 to 1,470]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Expert · high

#59 / 371
1,491[1,483 to 1,500]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · French · medium

#61 / 301
1,479[1,462 to 1,496]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Coding · high

#65 / 416
1,508[1,502 to 1,514]UnknownStandard11 Nov 2026
LMArena Text (style control)

LMArena · Industry software and IT services · medium

#65 / 421
1,497[1,491 to 1,502]UnknownStandard11 Nov 2026
LMArena WebDev

LMArena · Webdev HTML · medium

#65 / 155
1,522[1,505 to 1,539]UnknownStandard15 Nov 2026
LMArena Text (style control)

LMArena · Coding · medium

#67 / 416
1,506[1,500 to 1,512]UnknownStandard11 Nov 2026
LMArena WebDev

LMArena · Webdev HTML · high

#67 / 155
1,513[1,494 to 1,532]UnknownStandard15 Nov 2026
GPQA diamond

Epoch AI Benchmarking Hub · low

#68 / 338
88.9%[84.5% to 93.3%]UnknownStandard6 Aug 2026
LMArena Text (style control)

LMArena · Expert · medium

#68 / 371
1,487[1,478 to 1,496]UnknownStandard11 Nov 2026
ARC-AGI-2

Epoch AI Benchmarking Hub · high

#70 / 299
72.1%$0.85StandardNot stated
LMArena WebDev

LMArena · Overall · high

#70 / 156
1,500[1,493 to 1,507]UnknownStandard15 Nov 2026
LMArena WebDev

LMArena · Webdev · high

#70 / 156
1,500[1,493 to 1,507]UnknownStandard15 Nov 2026
LMArena WebDev

LMArena · Webdev react · high

#70 / 140
1,495[1,486 to 1,503]UnknownStandard15 Nov 2026
LMArena WebDev

LMArena · Overall · medium

#72 / 156
1,490[1,483 to 1,497]UnknownStandard15 Nov 2026
LMArena WebDev

LMArena · Webdev · medium

#72 / 156
1,490[1,483 to 1,497]UnknownStandard15 Nov 2026
LMArena WebDev

LMArena · Webdev react · medium

#72 / 140
1,480[1,473 to 1,488]UnknownStandard15 Nov 2026
GPQA diamond

Epoch AI Benchmarking Hub · minimal

#95 / 338
86.4%[81.6% to 91.2%]UnknownStandard15 Jul 2026
ARC-AGI-2

Epoch AI Benchmarking Hub · minimal

#186 / 299
8.9%$0.11On frontier. Evidence grade A of A to D: strongest.Not stated

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

    8.9

    $0.11

    high

    72.1

    $0.85

  • GPQA diamond

    Epoch AI Benchmarking Hub

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

    86.4

    low

    88.9

    high

    92.8

  • LMArena Document (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1458.7

    high

    1455.3

  • LMArena Document

    LMArena

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

    1460.6

    high

    1461.8

  • LMArena Text (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1518.3

    high

    1514.2

  • LMArena Text (style control)

    LMArena

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

    1505.9

    high

    1508.0

  • LMArena Text (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1469.2

    high

    1465.3

  • LMArena Text (style control)

    LMArena

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

    1477.1

    high

    1478.2

  • LMArena Text (style control)

    LMArena

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

    1486.8

    high

    1488.2

  • LMArena Text (style control)

    LMArena

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

    1486.9

    high

    1491.3

  • LMArena Text (style control)

    LMArena

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

    1479.2

    high

    1494.4

  • LMArena Text (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1488.7

    high

    1483.9

  • LMArena Text (style control)

    LMArena

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

    1495.4

    high

    1496.0

  • LMArena Text (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1490.3

    high

    1490.1

  • LMArena Text (style control)

    LMArena

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

    1462.8

    high

    1467.8

  • LMArena Text (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1457.2

    high

    1454.8

  • LMArena Text (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1481.8

    high

    1478.5

  • LMArena Text (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1496.5

    high

    1489.5

  • LMArena Text (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1488.9

    high

    1482.8

  • LMArena Text (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1488.3

    high

    1485.8

  • LMArena Text (style control)

    LMArena

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

    1496.8

    high

    1501.7

  • LMArena Text (style control)

    LMArena

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

    1472.0

    high

    1474.9

  • LMArena Text (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1466.3

    high

    1465.4

  • LMArena Text (style control)

    LMArena

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

    1459.0

    high

    1481.8

  • LMArena Text (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1446.3

    high

    1439.5

  • LMArena Text (style control)

    LMArena

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

    1480.3

    high

    1483.1

  • LMArena Text (style control)

    LMArena

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

    1482.0

    high

    1497.4

  • LMArena Text (style control)

    LMArena

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

    1478.3

    high

    1479.7

  • LMArena Text (style control)

    LMArena

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

    1467.7

    high

    1470.1

  • LMArena Text (style control)

    LMArena

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

    1476.0

    high

    1478.4

  • LMArena Text (style control)

    LMArena

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

    1496.2

    high

    1504.6

  • LMArena Text (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1491.9

    high

    1491.3

  • LMArena Text (style control)

    LMArena

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

    1471.1

    high

    1474.1

  • LMArena Text

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1527.1

    high

    1525.7

  • LMArena Text

    LMArena

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

    1487.5

    high

    1492.9

  • LMArena Text

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1470.6

    high

    1469.0

  • LMArena Text

    LMArena

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

    1475.6

    high

    1480.2

  • LMArena Text

    LMArena

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

    1486.0

    high

    1491.6

  • LMArena Text

    LMArena

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

    1485.3

    high

    1494.1

  • LMArena Text

    LMArena

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

    1473.1

    high

    1490.9

  • LMArena Text

    LMArena

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

    1490.7

    high

    1491.2

  • LMArena Text

    LMArena

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

    1485.3

    high

    1489.1

  • LMArena Text

    LMArena

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

    1478.5

    high

    1482.5

  • LMArena Text

    LMArena

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

    1458.7

    high

    1467.0

  • LMArena Text

    LMArena

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

    1459.0

    high

    1459.3

  • LMArena Text

    LMArena

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

    1485.5

    high

    1486.5

  • LMArena Text

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1495.4

    high

    1493.4

  • LMArena Text

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1494.5

    high

    1490.5

  • LMArena Text

    LMArena

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

    1483.0

    high

    1484.2

  • LMArena Text

    LMArena

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

    1485.9

    high

    1494.0

  • LMArena Text

    LMArena

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

    1472.7

    high

    1478.6

  • LMArena Text

    LMArena

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

    1464.9

    high

    1467.3

  • LMArena Text

    LMArena

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

    1461.4

    high

    1489.3

  • LMArena Text

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1455.5

    high

    1448.6

  • LMArena Text

    LMArena

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

    1477.0

    high

    1483.6

  • LMArena Text

    LMArena

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

    1487.1

    high

    1504.4

  • LMArena Text

    LMArena

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

    1476.5

    high

    1480.4

  • LMArena Text

    LMArena

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

    1472.2

    high

    1476.9

  • LMArena Text

    LMArena

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

    1477.8

    high

    1482.9

  • LMArena Text

    LMArena

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

    1498.4

    high

    1515.4

  • LMArena Text

    LMArena

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

    1492.5

    high

    1493.9

  • LMArena Text

    LMArena

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

    1477.5

    high

    1481.7

  • LMArena Vision (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1344.6

    high

    1341.2

  • LMArena Vision (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1274.8

    high

    1271.7

  • LMArena Vision (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1310.8

    high

    1307.5

  • LMArena Vision (style control)

    LMArena

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

    1273.5

    high

    1277.3

  • LMArena Vision (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1314.2

    high

    1307.1

  • LMArena Vision (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1295.7

    high

    1292.9

  • LMArena Vision (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1297.7

    high

    1290.7

  • LMArena Vision (style control)

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1281.5

    high

    1281.1

  • LMArena Vision

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1376.0

    high

    1373.0

  • LMArena Vision

    LMArena

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

    1298.9

    high

    1299.1

  • LMArena Vision

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1321.6

    high

    1321.1

  • LMArena Vision

    LMArena

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

    1297.6

    high

    1304.6

  • LMArena Vision

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1327.6

    high

    1322.7

  • LMArena Vision

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1322.8

    high

    1321.3

  • LMArena Vision

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1315.3

    high

    1310.8

  • LMArena Vision

    LMArena

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

    1305.5

    high

    1307.5

  • LMArena WebDev

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1505.4

    high

    1500.1

  • LMArena WebDev

    LMArena

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

    1490.0

    high

    1499.9

  • LMArena WebDev

    LMArena

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

    1490.0

    high

    1499.9

  • LMArena WebDev

    LMArena

    1 regression. Evidence grade D of A to D: weakest.
    medium

    1522.1

    high

    1512.9

  • LMArena WebDev

    LMArena

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

    1480.2

    high

    1494.8

Closest head-to-head rivals

Direct pairwise win rates for Gemini 3.5 Flash across shared benchmark boards.