# ARC-AGI-3 leaderboard

Reasoning benchmark

> Every figure below is reproduced as its upstream source published it: nothing is modelled, estimated, interpolated or converted. `Unavailable` means no source published the value — it is never a zero. Each value carries its evidence state and the date it was observed.

Page: https://modelchoice.dev/leaderboards/benchmarks/reasoning-arc-agi3

Abstraction and Reasoning Corpus for AGI v3. Every model the catalog carries a published ARC-AGI-3 value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | Reasoning |
| Measure | Agentic task completion under a capped evaluation budget |
| Tasks | Interactive game-like tasks with hidden rules |
| Difficulty | Frontier agentic reasoning |
| Published by | [ARC-AGI-3: A New Challenge for Frontier Agentic Intelligence](https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf) |

An interactive successor to ARC-AGI-2 that evaluates whether an AI agent can learn unfamiliar task mechanics through action and feedback.

## ARC-AGI-3 ranking

27 models with a published ARC-AGI-3 value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.

| Rank | Model | Provider | Agentic task completion under a capped evaluation budget | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [GPT-6.2 Astra](https://modelchoice.dev/models/gpt-6-2-astra) | OpenAI | 70.306138 | Observed 2026-11-21 · source benchlm:models |
| 2 | [Claude Fable 5.2](https://modelchoice.dev/models/claude-fable-5-2) | Anthropic | 65.66 | Observed 2026-11-21 · source benchlm:models |
| 3 | [Gemini 4 Carbon](https://modelchoice.dev/models/gemini-4-carbon) | Google | 63.855138 | Observed 2026-11-21 · source benchlm:models |
| 4 | [GPT-6 Astra](https://modelchoice.dev/models/gpt-6-astra) | OpenAI | 62.712802 | Observed 2026-11-21 · source benchlm:models |
| 5 | [GPT-6.2 Sol](https://modelchoice.dev/models/gpt-6-2-sol) | OpenAI | 54.926689 | Observed 2026-11-21 · source benchlm:models |
| 6 | [GPT-6.1 Sol](https://modelchoice.dev/models/gpt-6-1-sol) | OpenAI | 52.72904 | Observed 2026-11-21 · source benchlm:models |
| 7 | [Gemini 4 Argon](https://modelchoice.dev/models/gemini-4-argon) | Google | 46.265668 | Observed 2026-11-21 · source benchlm:models |
| 8 | [Claude Opus 5](https://modelchoice.dev/models/claude-opus-5) | Anthropic | 30.16 | Observed 2026-11-21 · source benchlm:models |
| 9 | [Gemini 3.8 Flash](https://modelchoice.dev/models/gemini-3-8-flash) | Google | 10.366279 | Observed 2026-11-21 · source benchlm:models |
| 10 | [GPT-5.6 Sol](https://modelchoice.dev/models/gpt-5-6-sol) | OpenAI | 7.78 | Observed 2026-11-21 · source benchlm:models |
| 11 | [GPT-6 Sol](https://modelchoice.dev/models/gpt-6-sol) | OpenAI | 4.623568 | Observed 2026-11-21 · source benchlm:models |
| 12 | [Grok 4.6](https://modelchoice.dev/models/grok-4-6) | xAI | 2.11 | Observed 2026-11-21 · source benchlm:models |
| 13 | [Grok 4.7](https://modelchoice.dev/models/grok-4-7) | xAI | 1.816253 | Observed 2026-11-21 · source benchlm:models |
| 14 | [Claude Opus 4.8](https://modelchoice.dev/models/claude-opus-4-8) | Anthropic | 1.52 | Observed 2026-11-21 · source benchlm:models |
| 15 | [GPT-5.6 Terra](https://modelchoice.dev/models/gpt-5-6-terra) | OpenAI | 0.8 | Observed 2026-11-21 · source benchlm:models |
| 16 | [Claude Opus 4.6](https://modelchoice.dev/models/claude-opus-4-6) | Anthropic | 0.51 | Observed 2026-11-21 · source benchlm:models |
| 17 | [Claude Haiku 5.5](https://modelchoice.dev/models/claude-haiku-5-5) | Anthropic | 0.451371 | Observed 2026-11-21 · source benchlm:models |
| 18 | [GPT-5.5](https://modelchoice.dev/models/gpt-5-5) | OpenAI | 0.43 | Observed 2026-11-21 · source benchlm:models |
| 19 | [Gemini 3.1 Pro](https://modelchoice.dev/models/gemini-3-1-pro) | Google | 0.42 | Observed 2026-11-21 · source benchlm:models |
| 20 | [Grok 4.5](https://modelchoice.dev/models/grok-4-5) | xAI | 0.3 | Observed 2026-11-21 · source benchlm:models |
| 21 | [Inkling 2 Small](https://modelchoice.dev/models/inkling-2-small) | Thinking Machines Lab | 0.29737 | Observed 2026-11-21 · source benchlm:models |
| 22 | [GPT-5.4](https://modelchoice.dev/models/gpt-5-4) | OpenAI | 0.21 | Observed 2026-11-21 · source benchlm:models |
| 23 | [Claude Opus 4.7](https://modelchoice.dev/models/claude-opus-4-7) | Anthropic | 0.18 | Observed 2026-11-21 · source benchlm:models |
| 23 | [Claude Opus 4.7 (Adaptive)](https://modelchoice.dev/models/claude-opus-4-7-max) | Anthropic | 0.18 | Last good 2026-11-18 · source benchlm:models |
| 23 | [GPT-5.6 Luna](https://modelchoice.dev/models/gpt-5-6-luna) | OpenAI | 0.18 | Observed 2026-11-21 · source benchlm:models |
| 26 | [GPT-6 Luna](https://modelchoice.dev/models/gpt-6-luna) | OpenAI | 0.10361 | Observed 2026-11-21 · source benchlm:models |
| 27 | [Grok 4.20](https://modelchoice.dev/models/grok-4-20-beta) | xAI | 0.09 | Observed 2026-11-21 · source benchlm:models |

Rows are ordered by the value the source published, highest first. The source does not state whether a higher value is the better result, so this document does not either: for a benchmark that measures a rate of failure, read the table from the bottom.
