priorsBuilder

OpenAI · released 2025-12-11

GPT-5.2

Measured on 67 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. #99xhigh of 58737.7%
  2. #173medium of 58726.7%
  3. #341none of 5878.0%

Leader: Opus 5.5 · 61.4%

knowledge

AA-Omniscience

526 results
  1. #130medium of 5260.3
  2. #140xhigh of 526-0.9
  3. #204none of 526-12.3

Leader: Opus 5.5 · 46.4

reasoning

BullshitBench v2

196 results
  1. #102none of 19638.0%
  2. #129high of 19628.0%
  3. #132medium of 19627.0%

Leader: Opus 4.8 · 95.0%

math

FrontierMath Tier 4 (v2)

70 results
  1. #34xhigh of 7031.7%

Leader: GPT-6.1 Sol · 100.0%

reasoning

SimpleBench

97 results
  1. #59high of 9745.8%

Leader: Opus 5.5 · 88.4%

All 67 benchmarks

show 62 more

agentic

Terminal-Bench 2.0

40 results
  1. #8xhigh of 4064.9%

Leader: GPT-5.5 · 84.7%

agentic

Vending-Bench 2

67 results
  1. #46xhigh of 67$3,591

Leader: GPT-6 Astra · $15,515

agentic

τ²-Bench Telecom

393 results
  1. #99xhigh of 39384.8%
  2. #143medium of 39374.3%
  3. #210none of 39346.5%

Leader: GLM 5.2 · 99.1%

agentic

τ³-Banking

206 results
  1. #140none of 20611.1%

Leader: Qwen 3.8 Max · 51.3%

coding

Arena WebDev

122 results
  1. #77xhigh of 1221416

Leader: Opus 5.5 · 1813

coding

LiveCodeBench

282 results
  1. #4medium of 28289.4%
  2. #6xhigh of 28288.9%
  3. #78none of 28266.9%

Leader: Gemini 3 Pro · 91.7%

coding

LiveCodeBench (Vals)

127 results
  1. #33xhigh of 12785.4%

Leader: Fable 5.1 · 90.5%

coding

SWE-bench Verified (Vals)

85 results
  1. #40xhigh of 8575.8%

Leader: Opus 5 · 97.0%

coding

Terminal-Bench 2.0 (Vals)

63 results
  1. #18high of 6351.7%

Leader: GPT-5.5 · 73.2%

coding

Terminal-Bench Hard-AA

387 results
  1. #27xhigh of 38747.0%
  2. #43medium of 38743.2%
  3. #105none of 38731.8%

Leader: GPT-5.6 Sol · 65.9%

coding

Vibe Code Bench v1.1

105 results
  1. #49xhigh of 10553.5%

Leader: Sonnet 5.5 · 92.4%

coding

WeirdML v2

127 results
  1. #27xhigh of 12772.2%
  2. #38medium of 12763.4%
  3. #63none of 12749.6%

Leader: GPT-6 Astra Pro · 93.6%

composite

AA Intelligence Index v4.3.2

610 results
  1. #116xhigh of 61030.4
  2. #145medium of 61026.5
  3. #259none of 61017.0

Leader: Opus 5.5 · 57.6

composite

Arena Text

220 results
  1. #30xhigh of 2201476
  2. #85high of 2201437

Leader: Gemini 4 · 1525

composite

Arena Text — Multi-Turn

220 results
  1. #15xhigh of 2201495
  2. #86high of 2201442

Leader: Gemini 4 · 1553

composite

Vals RSI Index v1.1

24 results
  1. #21xhigh of 2415.8%

Leader: Opus 5.5 · 37.3%

composite

Vals RSI Index v1.1: Harness Engineering: Judge

24 results
  1. #23xhigh of 243.9%

Leader: GPT-6 Sol · 22.8%

composite

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

24 results
  1. #10xhigh of 240.0%

Leader: Opus 5.5 · 35.1%

composite

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

24 results
  1. #18xhigh of 2428.7%

Leader: Opus 5.5 · 46.5%

composite

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

24 results
  1. #21xhigh of 2430.6%

Leader: Opus 5.5 · 53.0%

creative

MineBench

72 results
  1. #38xhigh of 721505

Leader: GPT-6 Astra Pro · 2281

creative

VoxelBench (image)

16 results
  1. #7xhigh of 161382

Leader: GPT-6.1 Sol · 2214

creative

VoxelBench (text)

55 results
  1. #35xhigh of 551437

Leader: GPT-6 Astra · 2732

economics

CorpFin (Vals)

117 results
  1. #26xhigh of 11765.9%

Leader: Opus 5 · 73.2%

economics

MortgageTax (Vals)

84 results
  1. #28xhigh of 8467.1%

Leader: Opus 5 · 72.1%

economics

TaxEval (Vals)

125 results
  1. #11xhigh of 12575.8%

Leader: Muse Spark 1.2 · 80.4%

instruction

IFBench

399 results
  1. #31xhigh of 39975.4%
  2. #102medium of 39965.2%
  3. #186none of 39947.4%

Leader: Grok 4.3 · 83.3%

knowledge

AA-Omniscience Accuracy

526 results
  1. #92xhigh of 52644.3%
  2. #133medium of 52638.3%
  3. #173none of 52630.9%

Leader: Fable 5.1 · 67.2%

knowledge

AA-Omniscience Non-hallucination

526 results
  1. #165medium of 52638.4%
  2. #170none of 52637.7%
  3. #286xhigh of 52618.8%

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

knowledge

Arena Search

28 results
  1. #10xhigh of 281207

Leader: GPT-5.6 Sol · 1257

knowledge

CaseLaw (Vals)

53 results
  1. #8xhigh of 5366.0%

Leader: Grok 4.3 · 79.3%

knowledge

Global-MMLU-Lite

118 results
  1. #15medium of 11889.8%
  2. #28none of 11886.8%

Leader: Gemini 3.1 Pro · 93.2%

knowledge

GPQA Diamond

539 results
  1. #62xhigh of 53990.3%
  2. #112medium of 53986.4%
  3. #282none of 53971.2%

Leader: GPT-6 Astra · 96.3%

knowledge

GPQA Diamond (Vals)

122 results
  1. #20xhigh of 12291.7%

Leader: Gemini 3.1 Pro · 95.5%

knowledge

MedCode

94 results
  1. #19xhigh of 9449.7%

Leader: Opus 5 · 63.6%

knowledge

MedQA (Vals)

80 results
  1. #19xhigh of 8094.1%

Leader: o1 · 96.5%

knowledge

MedScribe

96 results
  1. #35xhigh of 9684.4%

Leader: Opus 5.5 · 91.4%

knowledge

MMLU Pro (Vals)

122 results
  1. #45xhigh of 12286.2%

Leader: Fable 5.1 · 92.4%

knowledge

MMLU-Pro

285 results
  1. #10xhigh of 28587.4%
  2. #23medium of 28585.9%
  3. #74none of 28581.4%

Leader: Gemini 3 Pro · 89.8%

knowledge

MMMU Pro (Vals)

81 results
  1. #17xhigh of 8186.7%

Leader: Fable 5.1 · 90.6%

knowledge

MMMU-Pro

265 results
  1. #105medium of 26574.6%
  2. #167none of 26565.8%

Leader: Opus 5.5 · 87.7%

long-context

AA-LCR v1.1

523 results
  1. #37xhigh of 52382.7%
  2. #207medium of 52370.3%
  3. #319none of 52345.7%

Leader: Kimi K3 · 88.7%

long-context

Arena Document

43 results
  1. #40high of 431405
  2. #43xhigh of 431401

Leader: Opus 5 · 1516

math

AIME (Vals)

82 results
  1. #2xhigh of 8296.9%

Leader: Gemini 3.1 Pro · 98.1%

math

AIME 2025

228 results
  1. #1xhigh of 22899.0%
  2. #4medium of 22896.7%
  3. #115none of 22851.0%

Leader: this model · 99.0%

math

FrontierMath Tiers 1–3 (v2)

115 results
  1. #27xhigh of 11567.4%

Leader: GPT-6.1 Sol · 93.7%

math

MGSM (Vals)

62 results
  1. #4xhigh of 6294.0%

Leader: Opus 4.5 · 95.2%

other

Arena Vision

116 results
  1. #31xhigh of 1161278
  2. #58high of 1161243

Leader: Fable 5 · 1308

other

Poker Agent

17 results
  1. #1xhigh of 171132

Leader: this model · 1132

other

T3 Code usage share

85 results
  1. #32xhigh of 850.0%

Leader: Opus 5.5 · 29.1%

reasoning

ARC-AGI-1

215 results
  1. #91xhigh of 21586.2%
  2. #108high of 21578.7%
  3. #120medium of 21572.7%
  4. #147low of 21555.7%

Leader: Fable 5 · 98.5%

reasoning

ARC-AGI-2

216 results
  1. #93xhigh of 21652.9%
  2. #100high of 21643.3%
  3. #121medium of 21626.7%
  4. #140low of 2169.7%

Leader: GPT-6 Astra · 95.0%

reasoning

CritPt

529 results
  1. #105xhigh of 52911.6%
  2. #134medium of 5297.9%
  3. #252none of 5290.6%

Leader: Muse Wonder Full (Max) · 32.9%

reasoning

Furniture Assembly

31 results
  1. #16xhigh of 3138.3%

Leader: Opus 5.5 · 83.3%

reasoning

LegalBench

135 results
  1. #64xhigh of 13582.8%

Leader: Fable 5 · 88.6%

reasoning

SAGE

80 results
  1. #22xhigh of 8049.3%

Leader: Opus 4.7 · 56.1%

reliability

Tool-call error rate

171 results
  1. #18xhigh of 1710.2%

Leader: GPT-5.4 Pro · 0.0%

reliability

Uptime (7d)

181 results
  1. #47xhigh of 18199.9%

Leader: GPT-5 Pro · 100.0%

speed

Latency (Time to First Token)

351 results
  1. #21none of 3510.7s
  2. #339xhigh of 351144s

Leader: Gemini 2.5 Flash-Lite · 0.3s

speed

Output Speed (tokens/s)

351 results
  1. #200xhigh of 35183.0
  2. #227none of 35172.2

Leader: Celeris-1 · 1522.5

speed

Provider latency

177 results
  1. #143xhigh of 1772.9s

Leader: Qwen 2.5 7B · 0.2s

speed

Provider throughput

176 results
  1. #67xhigh of 17674.8

Leader: Muse Spark 1.1 · 204.3