priorsBuilder

Anthropic · released 2025-09-29

Sonnet 4.5

Measured on 69 benchmarks.

General frontier

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

Your mix’s benchmarks

5 of 8 measured

reasoning

Humanity's Last Exam

587 results
  1. #227default of 58717.8%
  2. #358none of 5877.2%

Leader: Opus 5.5 · 61.4%

knowledge

AA-Omniscience

526 results
  1. #133default of 526-0.1
  2. #182none of 526-9.4

Leader: Opus 5.5 · 46.4

reasoning

BullshitBench v2

196 results
  1. #12high of 19679.0%
  2. #21none of 19674.0%

Leader: Opus 4.8 · 95.0%

math

FrontierMath Tier 4 (v2)

70 results
  1. #67high of 702.4%

Leader: GPT-6.1 Sol · 100.0%

reasoning

SimpleBench

97 results
  1. #47high of 9754.3%

Leader: Opus 5.5 · 88.4%

All 69 benchmarks

show 64 more

agentic

AA-Briefcase v1.1

224 results
  1. #161default of 224713

Leader: Sonnet 5.5 · 1823

agentic

AutomationBench-AA

219 results
  1. #140default of 21914.1%

Leader: Gemini 4 · 77.5%

agentic

OSWorld-Verified

13 results
  1. #9high of 1362.9%

Leader: Fable 5 · 86.0%

agentic

Terminal-Bench 2.0

40 results
  1. #18high of 4046.5%

Leader: GPT-5.5 · 84.7%

agentic

Vending-Bench 2

67 results
  1. #42high of 67$3,839

Leader: GPT-6 Astra · $15,515

agentic

τ²-Bench Telecom

393 results
  1. #133default of 39378.1%
  2. #157none of 39370.5%

Leader: GLM 5.2 · 99.1%

agentic

τ³-Banking

206 results
  1. #85default of 20624.5%

Leader: Qwen 3.8 Max · 51.3%

coding

Arena WebDev

122 results
  1. #87high of 1221393

Leader: Opus 5.5 · 1813

coding

LiveCodeBench

282 results
  1. #56default of 28271.4%
  2. #105none of 28259.0%

Leader: Gemini 3 Pro · 91.7%

coding

LiveCodeBench (Vals)

127 results
  1. #81high of 12773.0%

Leader: Fable 5.1 · 90.5%

coding

SciCode

226 results
  1. #136default of 22645.7%

Leader: Opus 5.5 · 66.9%

coding

SWE-bench Verified (Vals)

85 results
  1. #59high of 8570.0%

Leader: Opus 5 · 97.0%

coding

Terminal-Bench 2.0 (Vals)

63 results
  1. #28high of 6341.6%

Leader: GPT-5.5 · 73.2%

coding

Terminal-Bench Hard-AA

387 results
  1. #79default of 38735.6%
  2. #121none of 38728.8%

Leader: GPT-5.6 Sol · 65.9%

coding

Terminal-Bench v2.1-AA

238 results
  1. #106default of 23855.8%

Leader: Fable 5.1 · 91.4%

coding

Terminal-Bench v4-AA

216 results
  1. #162default of 2160.0%

Leader: Sonnet 5.5 · 63.6%

coding

Vibe Code Bench v1.1

105 results
  1. #70high of 10522.6%

Leader: Sonnet 5.5 · 92.4%

composite

AA Intelligence Index v4.3.2

610 results
  1. #218default of 61020.7
  2. #236none of 61019.3

Leader: Opus 5.5 · 57.6

composite

Arena Text

220 results
  1. #57high of 2201456

Leader: Gemini 4 · 1525

composite

Arena Text — Multi-Turn

220 results
  1. #42high of 2201478

Leader: Gemini 4 · 1553

cost

AA Intelligence Index Cost per Task

177 results
  1. #95default of 177$0.56

Leader: GPT-6 Luna · $0.00

cost

Cost to Run AA Intelligence Index

177 results
  1. #117default of 177$1,501

Leader: GPT-6 Luna · $10.63

creative

MineBench

72 results
  1. #56high of 721186

Leader: GPT-6 Astra Pro · 2281

creative

VoxelBench (text)

55 results
  1. #51high of 551190

Leader: GPT-6 Astra · 2732

economics

CorpFin (Vals)

117 results
  1. #50high of 11762.0%

Leader: Opus 5 · 73.2%

economics

GDP.pdf

210 results
  1. #160default of 2105.2%

Leader: Muse Wonder Full (Max) · 32.6%

economics

GDPval-AA v2.1

288 results
  1. #175default of 288909

Leader: Opus 5.5 · 1866

economics

MortgageTax (Vals)

84 results
  1. #47high of 8464.0%

Leader: Opus 5 · 72.1%

economics

TaxEval (Vals)

125 results
  1. #49high of 12573.3%

Leader: Muse Spark 1.2 · 80.4%

instruction

IFBench

399 results
  1. #137default of 39957.3%
  2. #233none of 39942.7%

Leader: Grok 4.3 · 83.3%

knowledge

AA-Omniscience Accuracy

526 results
  1. #163default of 52632.9%
  2. #194none of 52628.4%

Leader: Fable 5.1 · 67.2%

knowledge

AA-Omniscience Non-hallucination

526 results
  1. #106default of 52650.9%
  2. #129none of 52647.3%

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

knowledge

Arena Search

28 results
  1. #21high of 281158

Leader: GPT-5.6 Sol · 1257

knowledge

CaseLaw (Vals)

53 results
  1. #19high of 5362.2%

Leader: Grok 4.3 · 79.3%

knowledge

Global-MMLU-Lite

118 results
  1. #20default of 11889.3%

Leader: Gemini 3.1 Pro · 93.2%

knowledge

GPQA Diamond

539 results
  1. #159default of 53983.4%
  2. #268none of 53972.7%

Leader: GPT-6 Astra · 96.3%

knowledge

GPQA Diamond (Vals)

122 results
  1. #63high of 12281.6%

Leader: Gemini 3.1 Pro · 95.5%

knowledge

MedCode

94 results
  1. #39high of 9444.1%

Leader: Opus 5 · 63.6%

knowledge

MedQA (Vals)

80 results
  1. #15high of 8094.7%

Leader: o1 · 96.5%

knowledge

MedScribe

96 results
  1. #32high of 9684.5%

Leader: Opus 5.5 · 91.4%

knowledge

MMLU Pro (Vals)

122 results
  1. #28high of 12287.4%

Leader: Fable 5.1 · 92.4%

knowledge

MMLU-Pro

285 results
  1. #9default of 28587.5%
  2. #20none of 28586.0%

Leader: Gemini 3 Pro · 89.8%

knowledge

MMMU Pro (Vals)

81 results
  1. #46high of 8179.3%

Leader: Fable 5.1 · 90.6%

knowledge

MMMU-Pro

265 results
  1. #157default of 26568.7%
  2. #171none of 26565.2%

Leader: Opus 5.5 · 87.7%

long-context

AA-LCR v1.1

523 results
  1. #186default of 52372.3%
  2. #291none of 52354.0%

Leader: Kimi K3 · 88.7%

long-context

Arena Document

43 results
  1. #28high of 431450

Leader: Opus 5 · 1516

math

AIME (Vals)

82 results
  1. #30high of 8288.2%

Leader: Gemini 3.1 Pro · 98.1%

math

AIME 2025

228 results
  1. #35default of 22888.0%
  2. #142none of 22837.0%

Leader: GPT-5.2 · 99.0%

math

FrontierMath Tiers 1–3 (v2)

115 results
  1. #80high of 11523.9%

Leader: GPT-6.1 Sol · 93.7%

math

MGSM (Vals)

62 results
  1. #3high of 6294.3%

Leader: Opus 4.5 · 95.2%

other

Poker Agent

17 results
  1. #8high of 171056

Leader: GPT-5.2 · 1132

other

T3 Code usage share

85 results
  1. #27high of 850.0%

Leader: Opus 5.5 · 29.1%

reasoning

ARC-AGI-1

215 results
  1. #133high of 21563.7%

Leader: Fable 5 · 98.5%

reasoning

ARC-AGI-2

216 results
  1. #132high of 21613.6%

Leader: GPT-6 Astra · 95.0%

reasoning

CritPt

529 results
  1. #220default of 5291.1%
  2. #319none of 5290.0%

Leader: Muse Wonder Full (Max) · 32.9%

reasoning

LegalBench

135 results
  1. #38high of 13584.1%

Leader: Fable 5 · 88.6%

reasoning

SAGE

80 results
  1. #63high of 8036.1%

Leader: Opus 4.7 · 56.1%

reasoning

SimpleQA Verified

47 results
  1. #30high of 4729.2%

Leader: Gemini 3.1 Pro · 77.5%

reliability

Tool-call error rate

171 results
  1. #116high of 1712.6%

Leader: GPT-5.4 Pro · 0.0%

reliability

Uptime (7d)

181 results
  1. #75high of 18199.7%

Leader: GPT-5 Pro · 100.0%

speed

Latency (Time to First Token)

351 results
  1. #56none of 3510.9s
  2. #274default of 35118s

Leader: Gemini 2.5 Flash-Lite · 0.3s

speed

Output Speed (tokens/s)

351 results
  1. #301default of 35147.4
  2. #319none of 35141.9

Leader: Celeris-1 · 1522.5

speed

Provider latency

177 results
  1. #94high of 1771.3s

Leader: Qwen 2.5 7B · 0.2s

speed

Provider throughput

176 results
  1. #146high of 17638.0

Leader: Muse Spark 1.1 · 204.3