priorsBuilder

Anthropic · released 2026-05-28

Opus 4.8

Measured on 103 benchmarks across 5 reasoning efforts.

General frontier

50.6

#14 of 32 models · best at max

Reasoning efforts in your mix

max50.6#47 of 128
xhigh50.5#48 of 128
high48.1#55 of 128
medium45.8#61 of 128
low39.0#88 of 128

Your mix’s benchmarks

8 of 8 measured

reasoning

Humanity's Last Exam

587 results
  1. #32max of 58748.7%

Leader: Opus 5.5 · 61.4%

knowledge

AA-Omniscience

526 results
  1. #36max of 52628.8

Leader: Opus 5.5 · 46.4

reasoning

ARC-AGI-3

57 results
  1. #28high of 571.5%

Leader: GPT-6 Astra · 99.9%

reasoning

BullshitBench v2

196 results
  1. #1none of 19695.0%
  2. #3xhigh of 19694.0%

Leader: this model · 95.0%

coding

DeepSWE v1.1

70 results
  1. #35max of 7059.0%
  2. #40xhigh of 7054.4%
  3. #48high of 7051.8%
  4. #51medium of 7048.7%
  5. #58low of 7040.8%

Leader: GPT-6 Astra · 74.1%

math

FrontierMath Tier 4 (v2)

70 results
  1. #22max of 7056.1%

Leader: GPT-6.1 Sol · 100.0%

reasoning

SimpleBench

97 results
  1. #25max of 9764.8%

Leader: Opus 5.5 · 88.4%

agentic

Terminal-Bench 4.0

35 results
  1. #26max of 3523.6%

Leader: Opus 5.5 · 64.8%

All 103 benchmarks

show 95 more

agentic

AA-AnalystAgent

44 results
  1. #14max of 4445.0%

Leader: Gemini 3.7 Flash · 60.0%

agentic

AA-Briefcase v1.1

224 results
  1. #68max of 2241322

Leader: Sonnet 5.5 · 1823

agentic

Agent Arena

50 results
  1. #12high of 507.1%

Leader: Opus 5.5 · 15.5%

agentic

AutomationBench (Zapier)

115 results
  1. #46xhigh of 11516.9%
  2. #49max of 11516.4%
  3. #50high of 11516.1%
  4. #56medium of 11514.0%
  5. #58low of 11513.6%

Leader: Gemini 4 · 51.3%

agentic

AutomationBench-AA

219 results
  1. #86max of 21945.6%

Leader: Gemini 4 · 77.5%

agentic

EnterpriseOps-Gym-AA

50 results
  1. #18max of 5044.0%

Leader: Fable 5 · 51.1%

agentic

Harvey's Legal Agent Benchmark

79 results
  1. #20max of 799.6%

Leader: Muse Spark 1.2 · 25.4%

agentic

Legal Research Bench

78 results
  1. #21max of 7843.8%

Leader: Muse Spark 1.3 · 55.3%

agentic

PostTrainBench v1.1

12 results
  1. #4max of 1233.8%
  2. #5xhigh of 1232.9%

Leader: Fable 5 · 41.8%

agentic

SkillsBench

34 results
  1. #8max of 3459.2%

Leader: DeepSeek V4.1 Flash · 69.8%

agentic

Tax Agent Bench

69 results
  1. #27max of 6930.0%

Leader: Fable 5.1 · 49.2%

agentic

Terminal-Bench 2.1

18 results
  1. #10high of 1878.9%

Leader: GPT-6 Astra · 87.4%

agentic

Terminal-Bench 4.0 (Vals)

47 results
  1. #21max of 4723.2%

Leader: Opus 5.5 · 65.2%

agentic

Terminal-Bench Science (Vals)

40 results
  1. #28max of 404.3%

Leader: GPT-6 Astra · 62.9%

agentic

Time Horizon Index: KSP

13 results
  1. #8max of 1311.8%

Leader: Opus 5.5 · 91.3%

agentic

Vending-Bench 2

67 results
  1. #22high of 67$5,787
  2. #48max of 67$2,992

Leader: GPT-6 Astra · $15,515

agentic

Vending-Bench Arena

12 results
  1. #9max of 12$5,188

Leader: GPT-6 Astra · $12,400

agentic

τ²-Bench Telecom

393 results
  1. #30max of 39394.4%

Leader: GLM 5.2 · 99.1%

agentic

τ³-Banking

206 results
  1. #53max of 20634.2%

Leader: Qwen 3.8 Max · 51.3%

coding

Arena WebDev

122 results
  1. #35high of 1221556
  2. #44max of 1221536

Leader: Opus 5.5 · 1813

coding

Code Migration

77 results
  1. #16max of 7747.3%

Leader: Sonnet 5.5 · 69.8%

coding

DeepSWE v1

29 results
  1. #3max of 2958.2%
  2. #4xhigh of 2957.7%
  3. #7high of 2950.7%
  4. #9medium of 2947.3%

Leader: GPT-5.5 · 70.0%

coding

Drone-Bench

11 results
  1. #6max of 1182.3%

Leader: Opus 5.5 · 97.3%

coding

FrontierCode 1.1 Extended

130 results
  1. #34max of 13059.6%
  2. #35xhigh of 13059.5%
  3. #61high of 13055.6%
  4. #62medium of 13055.4%
  5. #86low of 13050.8%

Leader: Opus 5.5 · 65.3%

coding

FrontierCode 1.1 Main

130 results
  1. #38max of 13046.5%
  2. #45xhigh of 13045.5%
  3. #66high of 13041.0%
  4. #68medium of 13040.5%
  5. #91low of 13035.1%

Leader: Opus 5.5 · 54.6%

coding

FrontierSWE v1

17 results
  1. #6max of 1766.5%

Leader: Fable 5 · 88.2%

coding

LiveCodeBench (Vals)

127 results
  1. #11max of 12787.8%

Leader: Fable 5.1 · 90.5%

coding

ProgramBench v1 Almost Resolved

62 results
  1. #18max of 6215.5%

Leader: Opus 5.5 · 65.0%

coding

ProgramBench v1 Fully Resolved

62 results
  1. #14max of 621.0%

Leader: Opus 5.5 · 18.5%

coding

ProgramBench v1 Raw Pass Rate

62 results
  1. #16max of 6271.9%

Leader: Opus 5.5 · 87.0%

coding

SciCode

226 results
  1. #60max of 22654.4%

Leader: Opus 5.5 · 66.9%

coding

SWE Atlas-QnA

22 results
  1. #4xhigh of 2257.3%

Leader: Opus 5 · 63.2%

coding

SWE-bench Verified (Vals)

85 results
  1. #12max of 8588.6%

Leader: Opus 5 · 97.0%

coding

Terminal-Bench 2.0 (Vals)

63 results
  1. #2max of 6370.0%

Leader: GPT-5.5 · 73.2%

coding

Terminal-Bench 2.1 (Vals)

74 results
  1. #25max of 7471.9%

Leader: Opus 5.5 · 87.6%

coding

Terminal-Bench Hard-AA

387 results
  1. #10max of 38758.3%

Leader: GPT-5.6 Sol · 65.9%

coding

Terminal-Bench v2.1-AA

238 results
  1. #29max of 23884.6%

Leader: Fable 5.1 · 91.4%

coding

Terminal-Bench v4-AA

216 results
  1. #56max of 21621.7%

Leader: Sonnet 5.5 · 63.6%

coding

Vibe Code Bench v1.1

105 results
  1. #20max of 10582.7%

Leader: Sonnet 5.5 · 92.4%

coding

WeirdML v2

127 results
  1. #16xhigh of 12782.9%
  2. #23medium of 12776.0%
  3. #29max of 12770.5%

Leader: GPT-6 Astra Pro · 93.6%

composite

AA Intelligence Index v4.3.2

610 results
  1. #54max of 61041.8

Leader: Opus 5.5 · 57.6

composite

Arena Text

220 results
  1. #24high of 2201481
  2. #39max of 2201474

Leader: Gemini 4 · 1525

composite

Arena Text — Multi-Turn

220 results
  1. #13high of 2201497
  2. #16max of 2201494

Leader: Gemini 4 · 1553

composite

Vals Index

47 results
  1. #15max of 4755.1%

Leader: Gemini 4 · 68.9%

composite

Vals Multimodal Index v1.2

33 results
  1. #5max of 3370.9%

Leader: Fable 5 · 74.2%

composite

Vals RSI Index v1.1

24 results
  1. #10max of 2424.0%

Leader: Opus 5.5 · 37.3%

composite

Vals RSI Index v1.1: Harness Engineering: Judge

24 results
  1. #9max of 2416.8%

Leader: GPT-6 Sol · 22.8%

composite

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

24 results
  1. #21max of 240.0%

Leader: Opus 5.5 · 35.1%

composite

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

24 results
  1. #10max of 2438.3%

Leader: Opus 5.5 · 46.5%

composite

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

24 results
  1. #11max of 2440.8%

Leader: Opus 5.5 · 53.0%

cost

AA Intelligence Index Cost per Task

177 results
  1. #167max of 177$4.08

Leader: GPT-6 Luna · $0.00

cost

Cost to Run AA Intelligence Index

177 results
  1. #170max of 177$6,874

Leader: GPT-6 Luna · $10.63

creative

MineBench

72 results
  1. #17max of 721820

Leader: GPT-6 Astra Pro · 2281

creative

VoxelBench (text)

55 results
  1. #28max of 551599

Leader: GPT-6 Astra · 2732

economics

CorpFin (Vals)

117 results
  1. #18max of 11766.7%

Leader: Opus 5 · 73.2%

economics

Excel Modeling Benchmark

74 results
  1. #14max of 7469.4%

Leader: Fable 5.1 · 76.7%

economics

Finance Agent v2

78 results
  1. #27max of 7853.9%

Leader: Gemini 4 · 65.4%

economics

GDP.pdf

210 results
  1. #45max of 21022.8%

Leader: Muse Wonder Full (Max) · 32.6%

economics

GDPval-AA v2.1

288 results
  1. #58max of 2881453

Leader: Opus 5.5 · 1866

economics

MortgageTax (Vals)

84 results
  1. #5max of 8469.9%

Leader: Opus 5 · 72.1%

economics

Public Benefits Bench v1 (Vals)

13 results
  1. #2max of 1362.1%

Leader: Fable 5 · 71.7%

economics

TaxEval (Vals)

125 results
  1. #14max of 12575.6%

Leader: Muse Spark 1.2 · 80.4%

instruction

IFBench

399 results
  1. #117max of 39962.2%

Leader: Grok 4.3 · 83.3%

knowledge

AA-Omniscience Accuracy

526 results
  1. #67max of 52648.8%

Leader: Fable 5.1 · 67.2%

knowledge

AA-Omniscience Non-hallucination

526 results
  1. #70max of 52660.7%

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

knowledge

Arena Search

28 results
  1. #11max of 281204

Leader: GPT-5.6 Sol · 1257

knowledge

GPQA Diamond

539 results
  1. #41max of 53992.0%

Leader: GPT-6 Astra · 96.3%

knowledge

GPQA Diamond (Vals)

122 results
  1. #16max of 12292.4%

Leader: Gemini 3.1 Pro · 95.5%

knowledge

MedCode

94 results
  1. #10max of 9453.2%

Leader: Opus 5 · 63.6%

knowledge

MedScribe

96 results
  1. #24max of 9685.8%

Leader: Opus 5.5 · 91.4%

knowledge

MMLU Pro (Vals)

122 results
  1. #9max of 12289.6%

Leader: Fable 5.1 · 92.4%

knowledge

MMMU Pro (Vals)

81 results
  1. #19max of 8186.6%

Leader: Fable 5.1 · 90.6%

knowledge

Public Benefits Bench v1.1

50 results
  1. #13max of 5068.1%

Leader: Opus 5 · 76.9%

long-context

AA-LCR v1.1

523 results
  1. #123max of 52377.7%

Leader: Kimi K3 · 88.7%

long-context

Arena Document

43 results
  1. #12high of 431475
  2. #19max of 431464

Leader: Opus 5 · 1516

long-context

MLCR-AA

94 results
  1. #14max of 9445.6%

Leader: Sonnet 5.5 · 75.0%

math

FrontierMath Tiers 1–3 (v2)

115 results
  1. #15max of 11580.0%

Leader: GPT-6.1 Sol · 93.7%

other

Arena Vision

116 results
  1. #20high of 1161286
  2. #29max of 1161279

Leader: Fable 5 · 1308

other

T3 Code usage share

85 results
  1. #13max of 850.5%

Leader: Opus 5.5 · 29.1%

reasoning

ARC-AGI-1

215 results
  1. #48max of 21592.5%
  2. #54high of 21592.0%
  3. #57medium of 21591.5%
  4. #79low of 21588.0%

Leader: Fable 5 · 98.5%

reasoning

ARC-AGI-2

216 results
  1. #48high of 21672.1%
  2. #50medium of 21671.7%
  3. #69low of 21662.2%

Leader: GPT-6 Astra · 95.0%

reasoning

Blueprint-Bench 2

28 results
  1. #26max of 2814.5%

Leader: Gemini 4 · 54.4%

reasoning

CritPt

529 results
  1. #55max of 52920.9%

Leader: Muse Wonder Full (Max) · 32.9%

reasoning

Furniture Assembly

31 results
  1. #13max of 3142.5%

Leader: Opus 5.5 · 83.3%

reasoning

Humanity's Sixth Sense

25 results
  1. #22max of 2523.5%

Leader: GPT-6 Astra · 53.6%

reasoning

LegalBench

135 results
  1. #52max of 13583.6%

Leader: Fable 5 · 88.6%

reasoning

Roboflow Visual Reasoning

102 results
  1. #76low of 10253.0%
  2. #78high of 10252.3%

Leader: GPT-6 Astra · 91.2%

reasoning

SAGE

80 results
  1. #3max of 8054.8%

Leader: Opus 4.7 · 56.1%

reasoning

SimpleQA Verified

47 results
  1. #19max of 4744.5%

Leader: Gemini 3.1 Pro · 77.5%

reliability

Tool-call error rate

171 results
  1. #78max of 1711.1%

Leader: GPT-5.4 Pro · 0.0%

reliability

Uptime (7d)

181 results
  1. #77max of 18199.7%

Leader: GPT-5 Pro · 100.0%

speed

Latency (Time to First Token)

351 results
  1. #318max of 35160s

Leader: Gemini 2.5 Flash-Lite · 0.3s

speed

Output Speed (tokens/s)

351 results
  1. #257max of 35160.4

Leader: Celeris-1 · 1522.5

speed

Provider latency

177 results
  1. #108max of 1771.6s

Leader: Qwen 2.5 7B · 0.2s

speed

Provider throughput

176 results
  1. #59max of 17680.0

Leader: Muse Spark 1.1 · 204.3