priorsBuilder

Google · released 2026-08-13

Gemini 3.7 Flash

Measured on 88 benchmarks across 3 reasoning efforts.

General frontier

44.9

#22 of 32 models · best at high

Reasoning efforts in your mix

high44.9#70 of 128
medium42.7#77 of 128
low38.8#90 of 128

Your mix’s benchmarks

6 of 8 measured

reasoning

Humanity's Last Exam

587 results
  1. #35high of 58747.9%
  2. #92medium of 58739.0%
  3. #115low of 58735.1%

Leader: Opus 5.5 · 61.4%

knowledge

AA-Omniscience

526 results
  1. #48high of 52626.5
  2. #53medium of 52623.7
  3. #56low of 52622.1

Leader: Opus 5.5 · 46.4

reasoning

BullshitBench v2

196 results
  1. #110minimal of 19635.0%
  2. #122max of 19631.0%

Leader: Opus 4.8 · 95.0%

coding

DeepSWE v1.1

70 results
  1. #25medium of 7065.5%
  2. #27high of 7065.3%
  3. #43low of 7053.8%

Leader: GPT-6 Astra · 74.1%

math

FrontierMath Tier 4 (v2)

70 results
  1. #29high of 7036.6%

Leader: GPT-6.1 Sol · 100.0%

agentic

Terminal-Bench 4.0

35 results
  1. #35high of 3511.2%

Leader: Opus 5.5 · 64.8%

All 88 benchmarks

show 82 more

agentic

AA-AnalystAgent

44 results
  1. #1high of 4460.0%

Leader: this model · 60.0%

agentic

AA-Briefcase v1.1

224 results
  1. #99high of 2241114

Leader: Sonnet 5.5 · 1823

agentic

Agent Arena

50 results
  1. #37high of 50-2.1%

Leader: Opus 5.5 · 15.5%

agentic

AutomationBench (Zapier)

115 results
  1. #15high of 11530.4%

Leader: Gemini 4 · 51.3%

agentic

AutomationBench-AA

219 results
  1. #25high of 21962.0%

Leader: Gemini 4 · 77.5%

agentic

EnterpriseOps-Gym-AA

50 results
  1. #2medium of 5050.4%

Leader: Fable 5 · 51.1%

agentic

Harvey's Legal Agent Benchmark

79 results
  1. #24high of 798.8%

Leader: Muse Spark 1.2 · 25.4%

agentic

Legal Research Bench

78 results
  1. #41high of 7834.6%

Leader: Muse Spark 1.3 · 55.3%

agentic

SkillsBench

34 results
  1. #3high of 3465.9%

Leader: DeepSeek V4.1 Flash · 69.8%

agentic

Tax Agent Bench

69 results
  1. #39high of 6922.8%

Leader: Fable 5.1 · 49.2%

agentic

Terminal-Bench 4.0 (Vals)

47 results
  1. #32high of 4712.1%

Leader: Opus 5.5 · 65.2%

agentic

Terminal-Bench Science (Vals)

40 results
  1. #19high of 407.1%

Leader: GPT-6 Astra · 62.9%

agentic

τ³-Banking

206 results
  1. #46medium of 20635.5%
  2. #56high of 20632.8%
  3. #69low of 20629.5%

Leader: Qwen 3.8 Max · 51.3%

coding

Arena Image-to-WebDev

58 results
  1. #13high of 581603

Leader: Opus 5.5 · 1749

coding

Arena WebDev

122 results
  1. #28high of 1221592

Leader: Opus 5.5 · 1813

coding

Code Migration

77 results
  1. #37high of 7734.8%

Leader: Sonnet 5.5 · 69.8%

coding

FrontierCode 1.1 Extended

130 results
  1. #54medium of 13056.3%
  2. #67high of 13054.7%
  3. #78low of 13051.8%

Leader: Opus 5.5 · 65.3%

coding

FrontierCode 1.1 Main

130 results
  1. #51medium of 13043.6%
  2. #58high of 13042.2%
  3. #82low of 13036.9%

Leader: Opus 5.5 · 54.6%

coding

FrontierSWE v2

21 results
  1. #15max of 2120.3%

Leader: GPT-6 Astra · 65.5%

coding

IOI

44 results
  1. #14high of 4467.8%

Leader: Gemini 4 · 100.0%

coding

LiveCodeBench (Vals)

127 results
  1. #5high of 12788.7%

Leader: Fable 5.1 · 90.5%

coding

ProgramBench v1 Almost Resolved

62 results
  1. #28high of 629.5%

Leader: Opus 5.5 · 65.0%

coding

ProgramBench v1 Fully Resolved

62 results
  1. #46high of 620.0%

Leader: Opus 5.5 · 18.5%

coding

ProgramBench v1 Raw Pass Rate

62 results
  1. #25high of 6268.7%

Leader: Opus 5.5 · 87.0%

coding

SciCode

226 results
  1. #12medium of 22659.8%
  2. #30high of 22657.2%
  3. #47low of 22655.7%

Leader: Opus 5.5 · 66.9%

coding

SWE-bench Verified (Vals)

85 results
  1. #22high of 8580.8%

Leader: Opus 5 · 97.0%

coding

Terminal-Bench 2.1 (Vals)

74 results
  1. #14high of 7477.5%

Leader: Opus 5.5 · 87.6%

coding

Terminal-Bench v2.1-AA

238 results
  1. #24high of 23885.8%
  2. #46low of 23879.8%
  3. #52medium of 23878.3%

Leader: Fable 5.1 · 91.4%

coding

Terminal-Bench v4-AA

216 results
  1. #73high of 21613.6%

Leader: Sonnet 5.5 · 63.6%

coding

Vibe Code Bench v1.1

105 results
  1. #38high of 10570.4%

Leader: Sonnet 5.5 · 92.4%

composite

AA Intelligence Index v4.3.2

610 results
  1. #64medium of 61039.6
  2. #69high of 61039.1
  3. #81low of 61036.9

Leader: Opus 5.5 · 57.6

composite

Arena Text

220 results
  1. #17high of 2201486

Leader: Gemini 4 · 1525

composite

Arena Text — Multi-Turn

220 results
  1. #19high of 2201492

Leader: Gemini 4 · 1553

composite

Vals Index

47 results
  1. #27high of 4751.3%

Leader: Gemini 4 · 68.9%

composite

Vals RSI Index v1.1

24 results
  1. #19max of 2418.3%

Leader: Opus 5.5 · 37.3%

composite

Vals RSI Index v1.1: Harness Engineering: Judge

24 results
  1. #20max of 248.4%

Leader: GPT-6 Sol · 22.8%

composite

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

24 results
  1. #14max of 240.0%

Leader: Opus 5.5 · 35.1%

composite

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

24 results
  1. #21max of 2425.0%

Leader: Opus 5.5 · 46.5%

composite

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

24 results
  1. #14max of 2439.7%

Leader: Opus 5.5 · 53.0%

cost

AA Intelligence Index Cost per Task

177 results
  1. #110high of 177$0.93

Leader: GPT-6 Luna · $0.00

cost

Cost to Run AA Intelligence Index

177 results
  1. #100high of 177$1,084

Leader: GPT-6 Luna · $10.63

creative

MineBench

72 results
  1. #18max of 721802

Leader: GPT-6 Astra Pro · 2281

creative

threejseval

34 results
  1. #16high of 341482

Leader: Opus 5.5 · 2079

creative

VoxelBench (text)

55 results
  1. #15max of 551832

Leader: GPT-6 Astra · 2732

economics

Excel Modeling Benchmark

74 results
  1. #12high of 7471.3%

Leader: Fable 5.1 · 76.7%

economics

Finance Agent v2

78 results
  1. #6high of 7859.0%

Leader: Gemini 4 · 65.4%

economics

GDP.pdf

210 results
  1. #42high of 21023.6%

Leader: Muse Wonder Full (Max) · 32.6%

economics

GDPval-AA v2.1

288 results
  1. #74high of 2881392
  2. #83medium of 2881358
  3. #93low of 2881324

Leader: Opus 5.5 · 1866

economics

MortgageTax (Vals)

84 results
  1. #30high of 8466.7%

Leader: Opus 5 · 72.1%

economics

TaxEval (Vals)

125 results
  1. #31high of 12574.7%

Leader: Muse Spark 1.2 · 80.4%

knowledge

AA-Omniscience Accuracy

526 results
  1. #36high of 52655.3%
  2. #41medium of 52654.0%
  3. #45low of 52653.6%

Leader: Fable 5.1 · 67.2%

knowledge

AA-Omniscience Non-hallucination

526 results
  1. #180high of 52635.5%
  2. #189medium of 52634.1%
  3. #204low of 52632.3%

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

knowledge

GPQA Diamond

539 results
  1. #6high of 53994.5%
  2. #39medium of 53992.1%
  3. #63low of 53990.1%

Leader: GPT-6 Astra · 96.3%

knowledge

GPQA Diamond (Vals)

122 results
  1. #5high of 12293.9%

Leader: Gemini 3.1 Pro · 95.5%

knowledge

MedCode

94 results
  1. #9high of 9453.4%

Leader: Opus 5 · 63.6%

knowledge

MedScribe

96 results
  1. #37high of 9683.9%

Leader: Opus 5.5 · 91.4%

knowledge

MMLU Pro (Vals)

122 results
  1. #6high of 12290.1%

Leader: Fable 5.1 · 92.4%

knowledge

MMMU Pro (Vals)

81 results
  1. #5high of 8189.0%

Leader: Fable 5.1 · 90.6%

knowledge

MMMU-Pro

265 results
  1. #10high of 26585.5%
  2. #14low of 26584.9%
  3. #16medium of 26584.7%

Leader: Opus 5.5 · 87.7%

long-context

AA-LCR v1.1

523 results
  1. #31medium of 52383.0%
  2. #50high of 52381.7%
  3. #108low of 52378.7%

Leader: Kimi K3 · 88.7%

long-context

MLCR-AA

94 results
  1. #49high of 9415.0%

Leader: Sonnet 5.5 · 75.0%

math

FrontierMath Tiers 1–3 (v2)

115 results
  1. #23high of 11571.6%

Leader: GPT-6.1 Sol · 93.7%

math

ProofBench v1.1

52 results
  1. #25high of 5258.0%

Leader: Sonnet 5.5 · 100.0%

other

Arena Vision

116 results
  1. #6high of 1161297

Leader: Fable 5 · 1308

other

T3 Code usage share

85 results
  1. #29max of 850.0%

Leader: Opus 5.5 · 29.1%

reasoning

ARC-AGI-1

215 results
  1. #30high of 21595.5%
  2. #59medium of 21591.2%
  3. #98low of 21585.2%

Leader: Fable 5 · 98.5%

reasoning

ARC-AGI-2

216 results
  1. #29high of 21684.6%
  2. #68medium of 21663.7%
  3. #92low of 21652.9%

Leader: GPT-6 Astra · 95.0%

reasoning

CritPt

529 results
  1. #94high of 52914.3%
  2. #117medium of 5299.4%
  3. #141low of 5295.7%

Leader: Muse Wonder Full (Max) · 32.9%

reasoning

Furniture Assembly

31 results
  1. #25high of 3126.7%

Leader: Opus 5.5 · 83.3%

reasoning

Humanity's Sixth Sense

25 results
  1. #6high of 2539.8%

Leader: GPT-6 Astra · 53.6%

reasoning

LegalBench

135 results
  1. #5high of 13587.3%

Leader: Fable 5 · 88.6%

reasoning

Roboflow Visual Reasoning

102 results
  1. #10high of 10281.9%
  2. #14low of 10280.8%

Leader: GPT-6 Astra · 91.2%

reasoning

SAGE

80 results
  1. #23high of 8049.2%

Leader: Opus 4.7 · 56.1%

reasoning

SimpleQA Verified

47 results
  1. #5max of 4773.2%

Leader: Gemini 3.1 Pro · 77.5%

reliability

Tool-call error rate

171 results
  1. #121max of 1712.9%

Leader: GPT-5.4 Pro · 0.0%

reliability

Uptime (7d)

181 results
  1. #137max of 18198.4%

Leader: GPT-5 Pro · 100.0%

safety

CyberBench v1.1

45 results
  1. #40high of 4543.8%

Leader: Wonder · 80.4%

safety

SRE Bench

32 results
  1. #15high of 324.6%

Leader: GPT-6 Astra · 56.9%

speed

Latency (Time to First Token)

351 results
  1. #94low of 3511.1s
  2. #232medium of 3515.9s
  3. #245high of 3518.5s

Leader: Gemini 2.5 Flash-Lite · 0.3s

speed

Output Speed (tokens/s)

351 results
  1. #8low of 351341.3
  2. #9medium of 351339.1
  3. #11high of 351317.9

Leader: Celeris-1 · 1522.5

speed

Provider latency

177 results
  1. #126max of 1772.2s

Leader: Qwen 2.5 7B · 0.2s

speed

Provider throughput

176 results
  1. #43max of 17686.5

Leader: Muse Spark 1.1 · 204.3