priorsBuilder

OpenAI · released 2026-03-05

GPT-5.4

Measured on 79 benchmarks.

General frontier

Not ranked in this mix — too little measured data for its benchmarks.

Your mix’s benchmarks

6 of 8 measured

reasoning

Humanity's Last Exam

587 results
  1. #53xhigh of 58743.7%
  2. #141low of 58730.8%
  3. #286none of 58711.3%

Leader: Opus 5.5 · 61.4%

knowledge

AA-Omniscience

526 results
  1. #100xhigh of 5265.8
  2. #105low of 5264.8
  3. #218none of 526-15.7

Leader: Opus 5.5 · 46.4

reasoning

ARC-AGI-3

57 results
  1. #47high of 570.2%

Leader: GPT-6 Astra · 99.9%

reasoning

BullshitBench v2

196 results
  1. #79none of 19648.0%
  2. #93xhigh of 19642.0%

Leader: Opus 4.8 · 95.0%

coding

DeepSWE v1.1

70 results
  1. #49xhigh of 7051.8%

Leader: GPT-6 Astra · 74.1%

math

FrontierMath Tier 4 (v2)

70 results
  1. #23xhigh of 7049.0%

Leader: GPT-6.1 Sol · 100.0%

All 79 benchmarks

show 73 more

agentic

Agent Arena

50 results
  1. #29high of 501.2%

Leader: Opus 5.5 · 15.5%

agentic

APEX-Agents-AA

33 results
  1. #9xhigh of 3333.3%

Leader: Gemini 3.5 Flash · 47.1%

agentic

AutomationBench (Zapier)

115 results
  1. #68xhigh of 11511.6%
  2. #71high of 11510.1%
  3. #90medium of 1155.3%
  4. #108low of 1151.5%

Leader: Gemini 4 · 51.3%

agentic

Harvey's Legal Agent Benchmark

79 results
  1. #76xhigh of 790.0%

Leader: Muse Spark 1.2 · 25.4%

agentic

PostTrainBench v1.1

12 results
  1. #12high of 1219.0%

Leader: Fable 5 · 41.8%

agentic

SkillsBench

34 results
  1. #18xhigh of 3451.7%

Leader: DeepSeek V4.1 Flash · 69.8%

agentic

Vending-Bench 2

67 results
  1. #20xhigh of 67$6,144

Leader: GPT-6 Astra · $15,515

agentic

τ²-Bench Telecom

393 results
  1. #80xhigh of 39387.1%
  2. #142low of 39374.6%
  3. #236none of 39336.0%

Leader: GLM 5.2 · 99.1%

agentic

τ³-Banking

206 results
  1. #31xhigh of 20639.6%

Leader: Qwen 3.8 Max · 51.3%

coding

Arena Image-to-WebDev

58 results
  1. #49xhigh of 581440

Leader: Opus 5.5 · 1749

coding

Arena WebDev

122 results
  1. #63high of 1221465
  2. #66medium of 1221444
  3. #82xhigh of 1221400

Leader: Opus 5.5 · 1813

coding

Code Migration

77 results
  1. #36xhigh of 7735.0%

Leader: Sonnet 5.5 · 69.8%

coding

DeepSWE v1

29 results
  1. #5xhigh of 2955.5%

Leader: GPT-5.5 · 70.0%

coding

FrontierSWE v1

17 results
  1. #10xhigh of 1746.0%

Leader: Fable 5 · 88.2%

coding

LiveCodeBench (Vals)

127 results
  1. #41xhigh of 12784.1%

Leader: Fable 5.1 · 90.5%

coding

ProgramBench v1 Almost Resolved

62 results
  1. #34xhigh of 627.0%
  2. #37high of 623.5%

Leader: Opus 5.5 · 65.0%

coding

ProgramBench v1 Fully Resolved

62 results
  1. #22high of 620.5%
  2. #54xhigh of 620.0%

Leader: Opus 5.5 · 18.5%

coding

ProgramBench v1 Raw Pass Rate

62 results
  1. #26xhigh of 6268.6%
  2. #35high of 6261.8%

Leader: Opus 5.5 · 87.0%

coding

SWE Atlas-QnA

22 results
  1. #10xhigh of 2240.8%

Leader: Opus 5 · 63.2%

coding

SWE-bench Verified (Vals)

85 results
  1. #29xhigh of 8578.2%

Leader: Opus 5 · 97.0%

coding

Terminal-Bench 2.0 (Vals)

63 results
  1. #12high of 6358.4%

Leader: GPT-5.5 · 73.2%

coding

Terminal-Bench Hard-AA

387 results
  1. #12xhigh of 38757.6%
  2. #42low of 38743.2%
  3. #68none of 38737.9%

Leader: GPT-5.6 Sol · 65.9%

coding

Terminal-Bench v2.1-AA

238 results
  1. #53xhigh of 23878.3%

Leader: Fable 5.1 · 91.4%

coding

Vibe Code Bench v1.1

105 results
  1. #41xhigh of 10567.4%

Leader: Sonnet 5.5 · 92.4%

coding

WeirdML v2

127 results
  1. #21xhigh of 12777.7%
  2. #49none of 12757.4%

Leader: GPT-6 Astra Pro · 93.6%

composite

AA Intelligence Index v4.3.2

610 results
  1. #70xhigh of 61039.0
  2. #137low of 61027.6
  3. #250none of 61018.2

Leader: Opus 5.5 · 57.6

composite

Arena Text

220 results
  1. #33high of 2201475
  2. #48xhigh of 2201465

Leader: Gemini 4 · 1525

composite

Arena Text — Multi-Turn

220 results
  1. #17high of 2201493
  2. #40xhigh of 2201480

Leader: Gemini 4 · 1553

composite

Vals RSI Index v1.1

24 results
  1. #22xhigh of 2415.5%

Leader: Opus 5.5 · 37.3%

composite

Vals RSI Index v1.1: Harness Engineering: Judge

24 results
  1. #19xhigh of 248.7%

Leader: GPT-6 Sol · 22.8%

composite

Vals RSI Index v1.1: Post-training: Finance Agent

24 results
  1. #19xhigh of 240.0%

Leader: Opus 5.5 · 35.1%

composite

Vals RSI Index v1.1: Pre-training: Compression

24 results
  1. #20xhigh of 2425.2%

Leader: Opus 5.5 · 46.5%

composite

Vals RSI Index v1.1: Pre-training: LM Training

24 results
  1. #22xhigh of 2428.0%

Leader: Opus 5.5 · 53.0%

creative

MineBench

72 results
  1. #25xhigh of 721701

Leader: GPT-6 Astra Pro · 2281

creative

VoxelBench (image)

16 results
  1. #4xhigh of 161442

Leader: GPT-6.1 Sol · 2214

creative

VoxelBench (text)

55 results
  1. #34xhigh of 551464

Leader: GPT-6 Astra · 2732

economics

CorpFin (Vals)

117 results
  1. #33xhigh of 11765.3%

Leader: Opus 5 · 73.2%

economics

GDPval-AA v2.1

288 results
  1. #107xhigh of 2881246

Leader: Opus 5.5 · 1866

economics

MortgageTax (Vals)

84 results
  1. #17xhigh of 8468.3%

Leader: Opus 5 · 72.1%

economics

TaxEval (Vals)

125 results
  1. #43xhigh of 12574.0%

Leader: Muse Spark 1.2 · 80.4%

instruction

IFBench

399 results
  1. #37xhigh of 39973.9%
  2. #100low of 39965.9%
  3. #180none of 39948.4%

Leader: Grok 4.3 · 83.3%

knowledge

AA-Omniscience Accuracy

526 results
  1. #59xhigh of 52650.8%
  2. #71low of 52647.9%
  3. #140none of 52637.4%

Leader: Fable 5.1 · 67.2%

knowledge

AA-Omniscience Non-hallucination

526 results
  1. #309low of 52617.4%
  2. #327none of 52615.3%
  3. #441xhigh of 5268.3%

Leader: MiniCPM5-1B (Non-reasoning) · 99.1%

knowledge

Arena Search

28 results
  1. #15xhigh of 281197

Leader: GPT-5.6 Sol · 1257

knowledge

CaseLaw (Vals)

53 results
  1. #16xhigh of 5363.8%

Leader: Grok 4.3 · 79.3%

knowledge

GPQA Diamond

539 results
  1. #42xhigh of 53992.0%
  2. #103low of 53987.1%
  3. #252none of 53974.8%

Leader: GPT-6 Astra · 96.3%

knowledge

GPQA Diamond (Vals)

122 results
  1. #21xhigh of 12291.7%

Leader: Gemini 3.1 Pro · 95.5%

knowledge

MedCode

94 results
  1. #51xhigh of 9441.3%

Leader: Opus 5 · 63.6%

knowledge

MedQA (Vals)

80 results
  1. #5xhigh of 8096.1%

Leader: o1 · 96.5%

knowledge

MedScribe

96 results
  1. #61xhigh of 9677.5%

Leader: Opus 5.5 · 91.4%

knowledge

MMLU Pro (Vals)

122 results
  1. #27xhigh of 12287.5%

Leader: Fable 5.1 · 92.4%

knowledge

MMMU Pro (Vals)

81 results
  1. #15xhigh of 8187.5%

Leader: Fable 5.1 · 90.6%

knowledge

MMMU-Pro

265 results
  1. #71xhigh of 26578.4%
  2. #74low of 26578.0%
  3. #138none of 26570.6%

Leader: Opus 5.5 · 87.7%

long-context

AA-LCR v1.1

523 results
  1. #45xhigh of 52382.0%
  2. #139low of 52376.7%
  3. #278none of 52358.3%

Leader: Kimi K3 · 88.7%

long-context

Arena Document

43 results
  1. #14xhigh of 431471

Leader: Opus 5 · 1516

math

AIME (Vals)

82 results
  1. #5xhigh of 8296.7%

Leader: Gemini 3.1 Pro · 98.1%

math

FrontierMath Tiers 1–3 (v2)

115 results
  1. #17xhigh of 11578.6%

Leader: GPT-6.1 Sol · 93.7%

other

Arena Vision

116 results
  1. #22high of 1161285
  2. #30xhigh of 1161279

Leader: Fable 5 · 1308

other

T3 Code usage share

85 results
  1. #23xhigh of 850.0%

Leader: Opus 5.5 · 29.1%

reasoning

ARC-AGI-1

215 results
  1. #42xhigh of 21593.7%
  2. #46high of 21592.7%
  3. #92medium of 21586.2%
  4. #127low of 21568.2%

Leader: Fable 5 · 98.5%

reasoning

ARC-AGI-2

216 results
  1. #46xhigh of 21674.0%
  2. #57high of 21667.5%
  3. #89medium of 21655.4%
  4. #118low of 21629.2%

Leader: GPT-6 Astra · 95.0%

reasoning

Blueprint-Bench 2

28 results
  1. #21xhigh of 2827.1%

Leader: Gemini 4 · 54.4%

reasoning

CritPt

529 results
  1. #48xhigh of 52923.4%
  2. #135low of 5297.4%
  3. #251none of 5290.6%

Leader: Muse Wonder Full (Max) · 32.9%

reasoning

Furniture Assembly

31 results
  1. #17xhigh of 3137.5%

Leader: Opus 5.5 · 83.3%

reasoning

LegalBench

135 results
  1. #16xhigh of 13586.0%

Leader: Fable 5 · 88.6%

reasoning

SAGE

80 results
  1. #48xhigh of 8043.3%

Leader: Opus 4.7 · 56.1%

reasoning

SimpleQA Verified

47 results
  1. #29xhigh of 4730.5%

Leader: Gemini 3.1 Pro · 77.5%

reliability

Tool-call error rate

171 results
  1. #84xhigh of 1711.3%

Leader: GPT-5.4 Pro · 0.0%

reliability

Uptime (7d)

181 results
  1. #154xhigh of 18197.7%

Leader: GPT-5 Pro · 100.0%

speed

Latency (Time to First Token)

351 results
  1. #44none of 3510.9s
  2. #144low of 3511.8s
  3. #345xhigh of 351197s

Leader: Gemini 2.5 Flash-Lite · 0.3s

speed

Output Speed (tokens/s)

351 results
  1. #184xhigh of 35191.6
  2. #251none of 35163.0
  3. #255low of 35161.3

Leader: Celeris-1 · 1522.5

speed

Provider latency

177 results
  1. #102xhigh of 1771.4s

Leader: Qwen 2.5 7B · 0.2s

speed

Provider throughput

176 results
  1. #98xhigh of 17656.6

Leader: Muse Spark 1.1 · 204.3