ARC-AGI-1 — leaderboard

ARC-AGI-1 evaluates model capability on agentic tasks from the linked upstream source with Score as the primary reported metric.

Metric: Accuracy (%). Source: arcprize.org. Status: saturated. 164 models tracked.

Top models

#ModelScore
1Gemini 3.1 Pro (Preview)98
2GPT-5.6 Sol (xHigh)97.5
3GPT-5.6 Sol (High)97
4GPT-5.6 Sol (Max)96.5
5GPT-5.6 Terra (Max)96.5
6Gemini 3 Deep Think96
7GPT-5.5 (xHigh)95
8GPT-5.5 Pro (xHigh)95
9GPT-5.5 (High)94.5
10GPT-5.4 Pro (xHigh)94.5
11GPT-5.6 Terra (xHigh)94
12GPT-5.4 (xHigh)93.67
13GPT-5.4 (High)92.67
14Gemini 3.5 Flash (High)92.5
15Claude Opus 4.8 (Max)92.5

Interactive version: theaggregate.ai/benchmark?slug=arc-agi-1 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.