Domains: MathScience
Skills: ReasoningMultimodal

Four-Factor Integrity Breakdown & Primary-Source Evidence

Privacy

Weight 40%
7.5/10
Access Model: Hybrid Public / Private
"Public Eval Set 120 tasks Calibrated, public, all tasks solved pass@2 by at least two humans, use to test your systems. Semi-Private Eval Set 120 tasks Calibrated, not public, all tasks solved pass@2 by at least two humans, used for Kaggle live contest leaderboard and ARC Prize leaderboard."
"ARC-AGI-2 contains 1,000 public training tasks and 120 public evaluation tasks."

Originality

Weight 25%
10/10
Reuse Profile: All Original (0% reuse)
"The Abstraction and Reasoning Corpus for Artificial General Intelligence (ARC-AGI), introduced in 2019, established a challenging benchmark for evaluating the general fluid intelligence of artificial systems via a set of unique, novel tasks only requiring minimal prior knowledge."
Primary source (Paper) ↗ Verified 2026-11-17

Freshness

Weight 20%
6/10
Initial Release: 2025-03Latest Refresh: 2025-03Rolling Updates: No
"By Greg Kamradt Published 24 Mar 2025 ARC-AGI-2 + ARC Prize 2025 is Live!"

Grading

Weight 15%
10/10
Integrity: Independent Offline (Leak-Verified)Sandbox: offline isolated
"Internet access is not available during Kaggle evaluation (no API-based systems like GPT/Claude/etc.)"

Frontier Model Standings on ARC-AGI-2

Top 15 active models tracked
Rank Model Official Score Percentile
#1GPT-6.2 Astra max OpenAI95.8%100.0p
#2GPT-6 Astra max OpenAI95.0%100.0p
#3GPT-6.1 Sol max OpenAI94.2%98.9p
#4Claude Fable 5.2 max Anthropic92.5%91.9p
#4Gemini 4 Hydrogen xhigh Google92.5%91.9p
#6Gemini 4 Carbon high Google92.1%85.1p
#7Claude Opus 5.5 max Anthropic91.7%80.5p
#8Gemini 4 Argon high Google91.2%75.9p
#9Claude Fable 5.1 max Anthropic90.0%71.3p
#10GPT-6 Sol max OpenAI89.6%66.7p
#11Gemini 3.8 Flash high Google89.2%62.1p
#12GPT-6.2 Sol max OpenAI85.8%57.5p
#13Gemini 3.7 Flash high Google84.6%53.0p
#14Inkling 2 Small default Thinking Machines81.7%48.4p
#15DeepSeek V4.1 Flash 1017 max DeepSeek78.1%43.7p

Embeddable Audit Badge

View Markdown (.md)

Benchmark maintainers can embed this live SVG badge in GitHub READMEs or documentation:

BenchAudit ARC-AGI-2 Tier A 8.2/10
[![BenchAudit Tier A](https://benchaudit.org/badges/arc-agi-2.svg)](https://benchaudit.org/benchmarks/arc-agi-2/)