priorsBuilder

Anthropic · released 2026-06-30

Sonnet 5

Measured on 90 benchmarks across 6 reasoning efforts.

General frontier

40.3

#28 of 32 models · best at max

Reasoning efforts in your mix

max40.3#84 of 128
xhigh36.8#98 of 128
high35.9#106 of 128
medium31.8#112 of 128
none30.0#116 of 128
low27.5#120 of 128

Your mix’s benchmarks

7 of 8 measured

reasoning

Humanity's Last Exam

587 results
  1. #74max of 58741.3%
  2. #91xhigh of 58739.0%
  3. #110high of 58735.7%
  4. #147medium of 58730.0%
  5. #197low of 58721.9%
  6. #217none of 58719.0%

Leader: Opus 5.5 · 61.4%

knowledge

AA-Omniscience

526 results
  1. #76max of 52616.4
  2. #114xhigh of 5263.2
  3. #135none of 526-0.7
  4. #150high of 526-3.7
  5. #168medium of 526-6.9
  6. #173low of 526-8.2

Leader: Opus 5.5 · 46.4

reasoning

BullshitBench v2

196 results
  1. #11low of 19680.0%
  2. #15max of 19678.0%

Leader: Opus 4.8 · 95.0%

coding

DeepSWE v1.1

70 results
  1. #42max of 7053.8%
  2. #50xhigh of 7049.7%
  3. #52high of 7048.2%
  4. #59medium of 7039.8%
  5. #64low of 7030.5%

Leader: GPT-6 Astra · 74.1%

math

FrontierMath Tier 4 (v2)

70 results
  1. #36max of 7029.3%

Leader: GPT-6.1 Sol · 100.0%

reasoning

SimpleBench

97 results
  1. #36max of 9760.6%

Leader: Opus 5.5 · 88.4%

agentic

Terminal-Bench 4.0

35 results
  1. #34max of 3512.4%

Leader: Opus 5.5 · 64.8%

All 90 benchmarks

show 83 more

agentic

AA-AnalystAgent

44 results
  1. #12max of 4446.3%

Leader: Gemini 3.7 Flash · 60.0%

agentic

AA-Briefcase v1.1

224 results
  1. #61max of 2241358
  2. #74xhigh of 2241273
  3. #91high of 2241176
  4. #110medium of 2241058
  5. #130low of 224927

Leader: Sonnet 5.5 · 1823

agentic

Agent Arena

50 results
  1. #14high of 504.9%

Leader: Opus 5.5 · 15.5%

agentic

AutomationBench (Zapier)

115 results
  1. #70max of 11510.7%
  2. #72high of 1159.9%
  3. #83xhigh of 1157.6%
  4. #88medium of 1155.8%
  5. #99low of 1153.6%

Leader: Gemini 4 · 51.3%

agentic

AutomationBench-AA

219 results
  1. #98max of 21936.5%
  2. #106xhigh of 21934.5%
  3. #110high of 21932.1%
  4. #118medium of 21927.9%
  5. #133low of 21919.9%

Leader: Gemini 4 · 77.5%

agentic

EnterpriseOps-Gym-AA

50 results
  1. #15max of 5044.7%
  2. #19xhigh of 5043.9%

Leader: Fable 5 · 51.1%

agentic

Harvey's Legal Agent Benchmark

79 results
  1. #37max of 795.0%

Leader: Muse Spark 1.2 · 25.4%

agentic

Legal Research Bench

78 results
  1. #23max of 7841.8%

Leader: Muse Spark 1.3 · 55.3%

agentic

SkillsBench

34 results
  1. #25max of 3446.5%

Leader: DeepSeek V4.1 Flash · 69.8%

agentic

Tax Agent Bench

69 results
  1. #28max of 6929.4%

Leader: Fable 5.1 · 49.2%

agentic

Terminal-Bench 2.1

18 results
  1. #14high of 1874.6%

Leader: GPT-6 Astra · 87.4%

agentic

Terminal-Bench 4.0 (Vals)

47 results
  1. #35max of 4710.6%

Leader: Opus 5.5 · 65.2%

agentic

Terminal-Bench Science (Vals)

40 results
  1. #22max of 405.7%

Leader: GPT-6 Astra · 62.9%

agentic

Vending-Bench 2

67 results
  1. #18max of 67$6,378

Leader: GPT-6 Astra · $15,515

agentic

τ³-Banking

206 results
  1. #41max of 20637.3%
  2. #115none of 20615.7%

Leader: Qwen 3.8 Max · 51.3%

coding

Arena Image-to-WebDev

58 results
  1. #31high of 581539

Leader: Opus 5.5 · 1749

coding

Arena WebDev

122 results
  1. #41high of 1221541

Leader: Opus 5.5 · 1813

coding

Code Migration

77 results
  1. #23max of 7744.4%

Leader: Sonnet 5.5 · 69.8%

coding

FrontierCode 1.1 Extended

130 results
  1. #55xhigh of 13056.2%
  2. #66max of 13054.9%
  3. #75high of 13052.7%
  4. #92medium of 13049.3%
  5. #101low of 13044.5%

Leader: Opus 5.5 · 65.3%

coding

FrontierCode 1.1 Main

130 results
  1. #54xhigh of 13042.7%
  2. #57max of 13042.4%
  3. #73high of 13039.4%
  4. #89medium of 13035.2%
  5. #102low of 13028.7%

Leader: Opus 5.5 · 54.6%

coding

IOI

44 results
  1. #32max of 4445.0%

Leader: Gemini 4 · 100.0%

coding

LiveCodeBench (Vals)

127 results
  1. #53max of 12782.4%

Leader: Fable 5.1 · 90.5%

coding

ProgramBench v1 Almost Resolved

62 results
  1. #22max of 6213.5%

Leader: Opus 5.5 · 65.0%

coding

ProgramBench v1 Fully Resolved

62 results
  1. #35max of 620.0%

Leader: Opus 5.5 · 18.5%

coding

ProgramBench v1 Raw Pass Rate

62 results
  1. #14max of 6272.1%

Leader: Opus 5.5 · 87.0%

coding

SciCode

226 results
  1. #61max of 22654.3%
  2. #62high of 22654.3%
  3. #69xhigh of 22654.1%
  4. #91medium of 22651.6%
  5. #107low of 22650.1%

Leader: Opus 5.5 · 66.9%

coding

SWE-bench Verified (Vals)

85 results
  1. #26max of 8579.6%

Leader: Opus 5 · 97.0%

coding

Terminal-Bench 2.1 (Vals)

74 results
  1. #19max of 7474.5%

Leader: Opus 5.5 · 87.6%

coding

Terminal-Bench v2.1-AA

238 results
  1. #42max of 23880.5%
  2. #62none of 23875.3%

Leader: Fable 5.1 · 91.4%

coding

Terminal-Bench v4-AA

216 results
  1. #71max of 21614.1%
  2. #92xhigh of 2167.1%
  3. #99high of 2165.1%
  4. #111low of 2162.5%
  5. #114medium of 2162.0%

Leader: Sonnet 5.5 · 63.6%

coding

Vibe Code Bench 1-100

22 results
  1. #18max of 2213.8%

Leader: Opus 5.5 · 30.4%

coding

Vibe Code Bench v1.1

105 results
  1. #23max of 10581.3%

Leader: Sonnet 5.5 · 92.4%

coding

WeirdML v2

127 results
  1. #32high of 12768.8%

Leader: GPT-6 Astra Pro · 93.6%

composite

AA Intelligence Index v4.3.2

610 results
  1. #74max of 61038.2
  2. #89xhigh of 61034.4
  3. #110high of 61031.7
  4. #134medium of 61028.1
  5. #173low of 61024.3
  6. #185none of 61023.2

Leader: Opus 5.5 · 57.6

composite

Arena Text

220 results
  1. #52high of 2201461

Leader: Gemini 4 · 1525

composite

Arena Text — Multi-Turn

220 results
  1. #48high of 2201473

Leader: Gemini 4 · 1553

composite

Vals Index

47 results
  1. #24max of 4751.9%

Leader: Gemini 4 · 68.9%

composite

Vals Multimodal Index v1.2

33 results
  1. #8max of 3368.8%

Leader: Fable 5 · 74.2%

cost

AA Intelligence Index Cost per Task

177 results
  1. #91low of 177$0.51
  2. #113medium of 177$1.00
  3. #143high of 177$1.79
  4. #159xhigh of 177$2.87
  5. #170max of 177$5.09

Leader: GPT-6 Luna · $0.00

cost

Cost to Run AA Intelligence Index

177 results
  1. #81low of 177$653
  2. #109medium of 177$1,218
  3. #136high of 177$2,086
  4. #152xhigh of 177$3,255
  5. #171max of 177$6,998

Leader: GPT-6 Luna · $10.63

creative

MineBench

72 results
  1. #27max of 721661

Leader: GPT-6 Astra Pro · 2281

creative

threejseval

34 results
  1. #29medium of 341219

Leader: Opus 5.5 · 2079

creative

VoxelBench (text)

55 results
  1. #32xhigh of 551554

Leader: GPT-6 Astra · 2732

economics

CorpFin (Vals)

117 results
  1. #15max of 11767.0%

Leader: Opus 5 · 73.2%

economics

Excel Modeling Benchmark

74 results
  1. #21max of 7466.3%

Leader: Fable 5.1 · 76.7%

economics

Finance Agent v2

78 results
  1. #28max of 7853.9%

Leader: Gemini 4 · 65.4%

economics

GDP.pdf

210 results
  1. #103max of 21013.2%
  2. #108xhigh of 21012.6%
  3. #113medium of 21011.6%
  4. #132high of 2109.4%
  5. #133low of 2109.4%

Leader: Muse Wonder Full (Max) · 32.6%

economics

GDPval-AA v2.1

288 results
  1. #55max of 2881463
  2. #84xhigh of 2881358
  3. #104high of 2881254
  4. #113none of 2881210
  5. #122medium of 2881157
  6. #146low of 2881066

Leader: Opus 5.5 · 1866

economics

MortgageTax (Vals)

84 results
  1. #4max of 8470.0%

Leader: Opus 5 · 72.1%

economics

TaxEval (Vals)

125 results
  1. #15max of 12575.6%

Leader: Muse Spark 1.2 · 80.4%

knowledge

AA-Omniscience Accuracy

526 results
  1. #122max of 52640.1%
  2. #128xhigh of 52639.0%
  3. #139high of 52637.4%
  4. #141low of 52637.4%
  5. #142medium of 52637.1%
  6. #160none of 52633.8%

Leader: Fable 5.1 · 67.2%

knowledge

AA-Omniscience Non-hallucination

526 results
  1. #71max of 52660.6%
  2. #126none of 52648.0%
  3. #149xhigh of 52641.4%
  4. #188high of 52634.3%
  5. #217medium of 52630.1%
  6. #224low of 52627.2%

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

knowledge

Arena Search

28 results
  1. #16max of 281194

Leader: GPT-5.6 Sol · 1257

knowledge

GPQA Diamond

539 results
  1. #50max of 53991.1%
  2. #191none of 53980.0%

Leader: GPT-6 Astra · 96.3%

knowledge

GPQA Diamond (Vals)

122 results
  1. #32max of 12288.9%

Leader: Gemini 3.1 Pro · 95.5%

knowledge

MedCode

94 results
  1. #31max of 9447.5%

Leader: Opus 5 · 63.6%

knowledge

MedScribe

96 results
  1. #69max of 9676.1%

Leader: Opus 5.5 · 91.4%

knowledge

MMLU Pro (Vals)

122 results
  1. #26max of 12287.5%

Leader: Fable 5.1 · 92.4%

knowledge

MMMU Pro (Vals)

81 results
  1. #35max of 8183.0%

Leader: Fable 5.1 · 90.6%

knowledge

MMMU-Pro

265 results
  1. #79max of 26577.3%
  2. #131none of 26571.9%

Leader: Opus 5.5 · 87.7%

knowledge

Public Benefits Bench v1.1

50 results
  1. #20max of 5066.0%

Leader: Opus 5 · 76.9%

long-context

AA-LCR v1.1

523 results
  1. #46max of 52382.0%
  2. #137xhigh of 52376.7%
  3. #138high of 52376.7%
  4. #172medium of 52373.7%
  5. #211none of 52370.0%
  6. #232low of 52367.3%

Leader: Kimi K3 · 88.7%

long-context

Arena Document

43 results
  1. #17high of 431466

Leader: Opus 5 · 1516

long-context

MLCR-AA

94 results
  1. #10max of 9455.0%

Leader: Sonnet 5.5 · 75.0%

math

FrontierMath Tiers 1–3 (v2)

115 results
  1. #32max of 11565.6%

Leader: GPT-6.1 Sol · 93.7%

math

ProofBench v1.1

52 results
  1. #17max of 5277.0%

Leader: Sonnet 5.5 · 100.0%

other

Arena Vision

116 results
  1. #42high of 1161263

Leader: Fable 5 · 1308

other

T3 Code usage share

85 results
  1. #8max of 851.6%

Leader: Opus 5.5 · 29.1%

reasoning

Blueprint-Bench 2

28 results
  1. #23max of 2824.9%

Leader: Gemini 4 · 54.4%

reasoning

CritPt

529 results
  1. #80max of 52916.9%
  2. #87xhigh of 52915.4%
  3. #90high of 52915.1%
  4. #128medium of 5298.6%
  5. #161low of 5294.6%
  6. #216none of 5291.1%

Leader: Muse Wonder Full (Max) · 32.9%

reasoning

Humanity's Sixth Sense

25 results
  1. #20max of 2524.8%

Leader: GPT-6 Astra · 53.6%

reasoning

LegalBench

135 results
  1. #45max of 13583.9%

Leader: Fable 5 · 88.6%

reasoning

Roboflow Visual Reasoning

102 results
  1. #85low of 10243.0%
  2. #86high of 10243.0%

Leader: GPT-6 Astra · 91.2%

reasoning

SAGE

80 results
  1. #24max of 8048.9%

Leader: Opus 4.7 · 56.1%

reasoning

SimpleQA Verified

47 results
  1. #26max of 4732.1%

Leader: Gemini 3.1 Pro · 77.5%

reliability

Tool-call error rate

171 results
  1. #112max of 1712.5%

Leader: GPT-5.4 Pro · 0.0%

reliability

Uptime (7d)

181 results
  1. #54max of 18199.9%

Leader: GPT-5 Pro · 100.0%

safety

CyberBench v1.1

45 results
  1. #22max of 4570.3%

Leader: Wonder · 80.4%

safety

SRE Bench

32 results
  1. #28max of 320.4%

Leader: GPT-6 Astra · 56.9%

speed

Latency (Time to First Token)

351 results
  1. #34none of 3510.8s
  2. #79low of 3511.0s
  3. #145medium of 3511.8s
  4. #258high of 35112s
  5. #305xhigh of 35138s
  6. #343max of 351171s

Leader: Gemini 2.5 Flash-Lite · 0.3s

speed

Output Speed (tokens/s)

351 results
  1. #204max of 35181.7
  2. #229xhigh of 35170.7
  3. #234high of 35168.7
  4. #240none of 35167.0
  5. #246medium of 35166.1
  6. #249low of 35165.1

Leader: Celeris-1 · 1522.5

speed

Provider latency

177 results
  1. #122max of 1771.9s

Leader: Qwen 2.5 7B · 0.2s

speed

Provider throughput

176 results
  1. #64max of 17675.3

Leader: Muse Spark 1.1 · 204.3