priorsBuilder

xAI · released 2026-03-09

Grok 4.20

Measured on 61 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. #121default of 58734.5%
  2. #166none of 58727.9%

Leader: Opus 5.5 · 61.4%

knowledge

AA-Omniscience

526 results
  1. #81default of 52614.8
  2. #327none of 526-42.6

Leader: Opus 5.5 · 46.4

reasoning

ARC-AGI-3

57 results
  1. #53xhigh of 570.1%

Leader: GPT-6 Astra · 99.9%

reasoning

BullshitBench v2

196 results
  1. #32low of 19667.0%
  2. #41xhigh of 19664.0%

Leader: Opus 4.8 · 95.0%

math

FrontierMath Tier 4 (v2)

70 results
  1. #53xhigh of 7017.1%

Leader: GPT-6.1 Sol · 100.0%

All 61 benchmarks

show 56 more

agentic

APEX-Agents-AA

33 results
  1. #26default of 3314.2%

Leader: Gemini 3.5 Flash · 47.1%

agentic

Harvey's Legal Agent Benchmark

79 results
  1. #73xhigh of 790.0%

Leader: Muse Spark 1.2 · 25.4%

agentic

Legal Research Bench

78 results
  1. #66xhigh of 7813.9%

Leader: Muse Spark 1.3 · 55.3%

agentic

Tax Agent Bench

69 results
  1. #61xhigh of 697.0%

Leader: Fable 5.1 · 49.2%

agentic

Terminal-Bench 2.0

40 results
  1. #13xhigh of 4057.3%

Leader: GPT-5.5 · 84.7%

agentic

Vending-Bench 2

67 results
  1. #35xhigh of 67$4,663

Leader: GPT-6 Astra · $15,515

agentic

τ²-Bench Telecom

393 results
  1. #13default of 39396.5%
  2. #159none of 39369.6%

Leader: GLM 5.2 · 99.1%

coding

Arena WebDev

122 results
  1. #92xhigh of 1221375

Leader: Opus 5.5 · 1813

coding

Code Migration

77 results
  1. #76xhigh of 770.3%

Leader: Sonnet 5.5 · 69.8%

coding

LiveCodeBench (Vals)

127 results
  1. #40xhigh of 12784.3%

Leader: Fable 5.1 · 90.5%

coding

SWE-bench Verified (Vals)

85 results
  1. #53xhigh of 8572.2%

Leader: Opus 5 · 97.0%

coding

Terminal-Bench 2.0 (Vals)

63 results
  1. #31xhigh of 6340.4%

Leader: GPT-5.5 · 73.2%

coding

Terminal-Bench 2.1 (Vals)

74 results
  1. #63xhigh of 7444.2%

Leader: Opus 5.5 · 87.6%

coding

Terminal-Bench Hard-AA

387 results
  1. #55default of 38740.9%
  2. #156none of 38722.0%

Leader: GPT-5.6 Sol · 65.9%

coding

Vibe Code Bench v1.1

105 results
  1. #93xhigh of 1054.1%

Leader: Sonnet 5.5 · 92.4%

coding

WeirdML v2

127 results
  1. #59xhigh of 12752.3%

Leader: GPT-6 Astra Pro · 93.6%

composite

AA Intelligence Index v4.3.2

610 results
  1. #155default of 61025.7
  2. #290none of 61014.6

Leader: Opus 5.5 · 57.6

composite

Arena Text

220 results
  1. #36xhigh of 2201475

Leader: Gemini 4 · 1525

composite

Arena Text — Multi-Turn

220 results
  1. #32xhigh of 2201483

Leader: Gemini 4 · 1553

composite

Vals Multimodal Index v1.2

33 results
  1. #32xhigh of 3339.1%

Leader: Fable 5 · 74.2%

creative

MineBench

72 results
  1. #46xhigh of 721329

Leader: GPT-6 Astra Pro · 2281

economics

CorpFin (Vals)

117 results
  1. #43xhigh of 11763.7%

Leader: Opus 5 · 73.2%

economics

Excel Modeling Benchmark

74 results
  1. #71xhigh of 7411.7%

Leader: Fable 5.1 · 76.7%

economics

Finance Agent v2

78 results
  1. #72xhigh of 7828.5%

Leader: Gemini 4 · 65.4%

economics

MortgageTax (Vals)

84 results
  1. #76xhigh of 8445.4%

Leader: Opus 5 · 72.1%

economics

TaxEval (Vals)

125 results
  1. #40xhigh of 12574.1%

Leader: Muse Spark 1.2 · 80.4%

instruction

IFBench

399 results
  1. #2default of 39982.9%
  2. #174none of 39949.3%

Leader: Grok 4.3 · 83.3%

knowledge

AA-Omniscience Accuracy

526 results
  1. #189default of 52628.9%
  2. #208none of 52626.7%

Leader: Fable 5.1 · 67.2%

knowledge

AA-Omniscience Non-hallucination

526 results
  1. #10default of 52682.6%
  2. #465none of 5267.2%

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

knowledge

Arena Search

28 results
  1. #12xhigh of 281204

Leader: GPT-5.6 Sol · 1257

knowledge

CaseLaw (Vals)

53 results
  1. #38xhigh of 5354.4%

Leader: Grok 4.3 · 79.3%

knowledge

Global-MMLU-Lite

118 results
  1. #18default of 11889.5%
  2. #50none of 11883.1%

Leader: Gemini 3.1 Pro · 93.2%

knowledge

GPQA Diamond

539 results
  1. #53default of 53991.1%
  2. #206none of 53978.5%

Leader: GPT-6 Astra · 96.3%

knowledge

GPQA Diamond (Vals)

122 results
  1. #34xhigh of 12288.6%

Leader: Gemini 3.1 Pro · 95.5%

knowledge

MedCode

94 results
  1. #82xhigh of 9432.2%

Leader: Opus 5 · 63.6%

knowledge

MedQA (Vals)

80 results
  1. #16xhigh of 8094.5%

Leader: o1 · 96.5%

knowledge

MedScribe

96 results
  1. #90xhigh of 9663.4%

Leader: Opus 5.5 · 91.4%

knowledge

MMLU Pro (Vals)

122 results
  1. #43xhigh of 12286.3%

Leader: Fable 5.1 · 92.4%

knowledge

MMMU Pro (Vals)

81 results
  1. #32xhigh of 8183.5%

Leader: Fable 5.1 · 90.6%

knowledge

MMMU-Pro

265 results
  1. #106default of 26574.6%
  2. #173none of 26564.9%

Leader: Opus 5.5 · 87.7%

long-context

AA-LCR v1.1

523 results
  1. #225default of 52369.0%
  2. #392none of 52325.3%

Leader: Kimi K3 · 88.7%

long-context

Arena Document

43 results
  1. #37xhigh of 431416

Leader: Opus 5 · 1516

math

AIME (Vals)

82 results
  1. #6xhigh of 8296.5%

Leader: Gemini 3.1 Pro · 98.1%

math

FrontierMath Tiers 1–3 (v2)

115 results
  1. #53xhigh of 11544.9%

Leader: GPT-6.1 Sol · 93.7%

other

Arena Vision

116 results
  1. #48xhigh of 1161256

Leader: Fable 5 · 1308

reasoning

ARC-AGI-1

215 results
  1. #74xhigh of 21589.5%

Leader: Fable 5 · 98.5%

reasoning

ARC-AGI-2

216 results
  1. #65xhigh of 21665.1%

Leader: GPT-6 Astra · 95.0%

reasoning

CritPt

529 results
  1. #139default of 5296.6%
  2. #280none of 5290.3%

Leader: Muse Wonder Full (Max) · 32.9%

reasoning

LegalBench

135 results
  1. #97xhigh of 13577.7%

Leader: Fable 5 · 88.6%

reasoning

SAGE

80 results
  1. #59xhigh of 8038.2%

Leader: Opus 4.7 · 56.1%

reliability

Tool-call error rate

171 results
  1. #6xhigh of 1710.0%

Leader: GPT-5.4 Pro · 0.0%

reliability

Uptime (7d)

181 results
  1. #89xhigh of 18199.6%

Leader: GPT-5 Pro · 100.0%

speed

Latency (Time to First Token)

351 results
  1. #12none of 3510.6s
  2. #281default of 35120s

Leader: Gemini 2.5 Flash-Lite · 0.3s

speed

Output Speed (tokens/s)

351 results
  1. #171default of 35195.4
  2. #183none of 35191.6

Leader: Celeris-1 · 1522.5

speed

Provider latency

177 results
  1. #32xhigh of 1770.6s

Leader: Qwen 2.5 7B · 0.2s

speed

Provider throughput

176 results
  1. #71xhigh of 17673.3

Leader: Muse Spark 1.1 · 204.3