Skip to content
Yardstick
  1. Benchmarks
  2. ARC-AGI-2

ARC-AGI-2

Visual grid puzzles that require inferring a brand-new rule from a handful of examples.

ReasoningRun by ARC Prize FoundationOfficial leaderboard (opens in a new tab)

Top score
95%GPT-6 Astra
Median of field
72.9%11 models scored
Human baseline
100%Human panel: every task solved by at least two people
Data as of
Oct 8, 2026ARC Prize Verified leaderboard

Standings

Bars share one axis from 0% to 100%.

  1. Rank 1
    GPT-6 Astra95%

    OpenAI · $10 / $50New

  2. Rank 2
    GPT-6.1 Sol94.2%

    OpenAI · $2 / $10New

  3. Rank 3
    Claude Opus 5.593.3%

    Anthropic · $4 / $20New

  4. Rank 4
    Claude Fable 5.190%

    Anthropic · $10 / $50New

  5. Rank 5
    Gemini 3.8 Flash89.2%

    Google · $0.75 / $3.75New

  6. Rank 6
    DeepSeek-V4.1-Flash72.9%

    DeepSeek · $0.30 / $1.20NewOpen weights

  7. Rank 7
    GLM-5.3-Flash65.8%

    Z.ai · $0.15 / $0.50NewOpen weights

  8. Rank 8
    Grok 4.761.4%

    SpaceXAI · $2 / $6New

  9. Rank 9
    Kimi K360.4%

    Moonshot AI · $3 / $15Open weights

  10. Rank 10
    GPT-6 Luna59.3%

    OpenAI · $0.10 / $0.50New

  11. Rank 11
    Qwen3.8-27B42.4%

    Alibaba · $0.50 / $3NewOpen weights

Not reported for ARC-AGI-2 · 18

  • ·
    Claude Haiku 5.5—Not reported

    Anthropic · $0.10 / $0.50New

  • ·
    Claude Sonnet 5.5—Not reported

    Anthropic · $2 / $10New

  • ·
    DeepSeek-V4-Pro—Not reported

    DeepSeek · $1.32 / $3.96NewOpen weights

  • ·
    Gemini 3.1 Pro—Not reported

    Google · $2 / $12Preview

  • ·
    Gemini 3.5 Flash-Lite—Not reported

    Google · $0.30 / $2.50

  • ·
    Gemini 4 Argon—Not reported

    Google · $2 / $10NewPreview

  • ·
    Gemma 4 31B—Not reported

    Google · Self-hostedOpen weights

  • ·
    GLM-5.3—Not reported

    Z.ai · $1.40 / $4.40NewOpen weights

  • ·
    Kimi K2.7 Code—Not reported

    Moonshot AI · $0.95 / $4Open weights

  • ·
    MiMo-V2.6-Pro—Not reported

    Xiaomi · $0.435 / $0.87NewOpen weights

  • ·
    MiniMax M3—Not reported

    MiniMax · $0.30 / $1.20Open weights

  • ·
    Mistral Large 4—Not reported

    Mistral AI · $1.36 / $4.18NewPreview

  • ·
    Mistral Medium 3.5—Not reported

    Mistral AI · $1.50 / $7.50Open weights

  • ·
    Mistral Small 4—Not reported

    Mistral AI · $0.15 / $0.60Open weights

  • ·
    Muse Glimmer 30B—Not reported

    Meta · Self-hostedNewOpen weights

  • ·
    Muse Spark 1.3—Not reported

    Meta · $1.25 / $4.25New

  • ·
    Qwen3.8-Flash-Next—Not reported

    Alibaba · $0.15 / $0.47NewOpen weights

  • ·
    Qwen3.8-Max—Not reported

    Alibaba · $2 / $6

What it measures

Fluid reasoning: working out an unfamiliar transformation rule from a few examples and applying it exactly. The tasks are easy for people and cannot be solved by recall.

How it is scored

Percent of 120 semi-private tasks solved with two attempts each, run by the ARC Prize Foundation itself, which also reports cost per task.

Read it with care

Close to its ceiling for frontier models — the best verified score is 95%. It still separates smaller and cheaper models, and cost per task varies more than tenfold between similar scores.

Other benchmarks

  • Artificial Analysis Intelligence Index
  • Epoch Capabilities Index
  • LiveBench
  • Terminal-Bench 4.0
  • SWE-Bench Pro (V2 Hard)
  • Arena WebDev
  • GDPval-AA
  • Vending-Bench 2
  • ARC-AGI-3
  • FrontierMath (Tiers 1–3)
  • MathArena
  • SimpleBench
  • Humanity's Last Exam (Diamond)
  • GPQA Diamond
  • Arena Vision
  • Arena Text
Yardstick

A plain reading of how today’s SI models compare. Scores are taken from each benchmark’s own leaderboard wherever one exists; anything a lab reported about its own model is marked.

Data as of Oct 8, 2026.

Explore

  • Models
  • Compare
  • Benchmarks
  • Value
  • Labs
  • Methodology & sources

Sources

  • Artificial Analysis (opens in a new tab)
  • Arena (opens in a new tab)
  • Epoch AI Benchmarking Hub (opens in a new tab)
  • ARC Prize (opens in a new tab)
  • Terminal-Bench (opens in a new tab)
  • Scale AI Labs (opens in a new tab)

Yardstick is an independent reference and is not affiliated with, sponsored by or endorsed by any lab listed. Product names and logos are trademarks of their respective owners and are used only to identify them. Benchmarks are imperfect proxies: check a model on your own task before you commit to it.

  • Models
  • Compare
  • Benchmarks
  • Value
  • Labs
  • Models
  • Compare
  • Benchmarks
  • Value
  • Labs
  • Models
  • Compare
  • Benchmarks
  • Value
  • Labs