# ARC-AGI-1 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://modelscale.dev/leaderboards/benchmarks/reasoning-arc-agi1

ARC-AGI-1 Semi-Private Evaluation. Every model the catalog carries a published ARC-AGI-1 value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | Reasoning |
| Measure | Verified accuracy |
| Tasks | Semi-private ARC-AGI-1 evaluation set |
| Difficulty | Abstract visual reasoning |
| Published by | [ARC Prize leaderboard](https://arcprize.org/) |

ARC Prize fluid-intelligence benchmark using novel visual grid transformations.

## ARC-AGI-1 ranking

43 models with a published ARC-AGI-1 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 | Verified accuracy | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [Claude Fable 5](https://modelscale.dev/models/claude-fable-5) | Anthropic | 98.50 | Observed 2026-11-21 · source benchlm:models |
| 1 | [Gemini 3.8 Flash](https://modelscale.dev/models/gemini-3-8-flash) | Google | 98.50 | Observed 2026-11-21 · source benchlm:models |
| 1 | [Gemini 4 Argon](https://modelscale.dev/models/gemini-4-argon) | Google | 98.50 | Observed 2026-11-21 · source benchlm:models |
| 1 | [GPT-6.2 Astra](https://modelscale.dev/models/gpt-6-2-astra) | OpenAI | 98.50 | Observed 2026-11-21 · source benchlm:models |
| 5 | [Claude Fable 5.1](https://modelscale.dev/models/claude-fable-5-1) | Anthropic | 97.50 | Observed 2026-11-21 · source benchlm:models |
| 5 | [Claude Opus 5](https://modelscale.dev/models/claude-opus-5) | Anthropic | 97.50 | Observed 2026-11-21 · source benchlm:models |
| 5 | [Claude Opus 5.5](https://modelscale.dev/models/claude-opus-5-5) | Anthropic | 97.50 | Observed 2026-11-21 · source benchlm:models |
| 5 | [Gemini 4 Carbon](https://modelscale.dev/models/gemini-4-carbon) | Google | 97.50 | Observed 2026-11-21 · source benchlm:models |
| 5 | [GPT-6 Astra](https://modelscale.dev/models/gpt-6-astra) | OpenAI | 97.50 | Observed 2026-11-21 · source benchlm:models |
| 5 | [GPT-6.2 Sol](https://modelscale.dev/models/gpt-6-2-sol) | OpenAI | 97.50 | Observed 2026-11-21 · source benchlm:models |
| 11 | [GPT-5.6 Sol](https://modelscale.dev/models/gpt-5-6-sol) | OpenAI | 96.50 | Observed 2026-11-21 · source benchlm:models |
| 11 | [GPT-5.6 Terra](https://modelscale.dev/models/gpt-5-6-terra) | OpenAI | 96.50 | Observed 2026-11-21 · source benchlm:models |
| 11 | [GPT-6.1 Sol](https://modelscale.dev/models/gpt-6-1-sol) | OpenAI | 96.50 | Observed 2026-11-21 · source benchlm:models |
| 11 | [Inkling 2 Small](https://modelscale.dev/models/inkling-2-small) | Thinking Machines Lab | 96.50 | Observed 2026-11-21 · source benchlm:models |
| 15 | [Gemini 3.7 Flash](https://modelscale.dev/models/gemini-3-7-flash) | Google | 95.50 | Observed 2026-11-21 · source benchlm:models |
| 15 | [GPT-6 Sol](https://modelscale.dev/models/gpt-6-sol) | OpenAI | 95.50 | Observed 2026-11-21 · source benchlm:models |
| 17 | [GPT-5.5](https://modelscale.dev/models/gpt-5-5) | OpenAI | 95.00 | Observed 2026-11-21 · source benchlm:models |
| 17 | [GPT-5.5 Pro](https://modelscale.dev/models/gpt-5-5-pro) | OpenAI | 95.00 | Observed 2026-11-21 · source benchlm:models |
| 19 | [GPT-5.4 Pro](https://modelscale.dev/models/gpt-5-4-pro) | OpenAI | 94.50 | Observed 2026-11-21 · source benchlm:models |
| 19 | [Kimi K3](https://modelscale.dev/models/kimi-3) | Moonshot AI | 94.50 | Observed 2026-11-21 · source benchlm:models |
| 21 | [GPT-5.4](https://modelscale.dev/models/gpt-5-4) | OpenAI | 93.67 | Observed 2026-11-21 · source benchlm:models |
| 22 | [Claude Opus 4.6 (Adaptive)](https://modelscale.dev/models/claude-opus-4-6-thinking) | Anthropic | 93.00 | Observed 2026-11-21 · source benchlm:models |
| 23 | [Claude Opus 4.8](https://modelscale.dev/models/claude-opus-4-8) | Anthropic | 92.50 | Observed 2026-11-21 · source benchlm:models |
| 23 | [Gemini 3.5 Flash](https://modelscale.dev/models/gemini-3-5-flash) | Google | 92.50 | Observed 2026-11-21 · source benchlm:models |
| 25 | [Gemini 3.6 Flash](https://modelscale.dev/models/gemini-3-6-flash) | Google | 91.17 | Observed 2026-11-21 · source benchlm:models |
| 26 | [DeepSeek V4 Pro](https://modelscale.dev/models/deepseek-v4-pro) | DeepSeek | 90.00 | Observed 2026-11-21 · source benchlm:models |
| 26 | [DeepSeek V4 Pro 0813](https://modelscale.dev/models/deepseek-v4-pro-max) | DeepSeek | 90.00 | Last good 2026-11-18 · source benchlm:models |
| 28 | [Grok 4.7](https://modelscale.dev/models/grok-4-7) | xAI | 89.50 | Observed 2026-11-21 · source benchlm:models |
| 29 | [Claude Haiku 5.5](https://modelscale.dev/models/claude-haiku-5-5) | Anthropic | 89.00 | Observed 2026-11-21 · source benchlm:models |
| 29 | [DeepSeek V4 Flash 0731](https://modelscale.dev/models/deepseek-v4-flash-max) | DeepSeek | 89.00 | Observed 2026-11-21 · source benchlm:models |
| 31 | [GPT-5.6 Luna](https://modelscale.dev/models/gpt-5-6-luna) | OpenAI | 88.00 | Observed 2026-11-21 · source benchlm:models |
| 32 | [Grok 4.6](https://modelscale.dev/models/grok-4-6) | xAI | 87.00 | Observed 2026-11-21 · source benchlm:models |
| 33 | [GPT-6 Luna](https://modelscale.dev/models/gpt-6-luna) | OpenAI | 86.67 | Observed 2026-11-21 · source benchlm:models |
| 34 | [GPT-5.2](https://modelscale.dev/models/gpt-5-2) | OpenAI | 86.17 | Observed 2026-11-21 · source benchlm:models |
| 35 | [Claude Sonnet 4.6](https://modelscale.dev/models/claude-sonnet-4-6) | Anthropic | 86.00 | Observed 2026-11-21 · source benchlm:models |
| 36 | [Grok 4.5](https://modelscale.dev/models/grok-4-5) | xAI | 85.67 | Observed 2026-11-21 · source benchlm:models |
| 37 | [Inkling-Small](https://modelscale.dev/models/inkling-small) | Thinking Machines Lab | 84.00 | Observed 2026-11-21 · source benchlm:models |
| 38 | [Claude Opus 4.5 Thinking](https://modelscale.dev/models/claude-opus-4-5-thinking) | Anthropic | 80.00 | Observed 2026-11-21 · source benchlm:models |
| 39 | [Gemini 3 Pro](https://modelscale.dev/models/gemini-3-pro) | Google | 75.00 | Observed 2026-11-21 · source benchlm:models |
| 40 | [GPT-5.1](https://modelscale.dev/models/gpt-5-1) | OpenAI | 72.83 | Observed 2026-11-21 · source benchlm:models |
| 41 | [GPT-5.4 mini](https://modelscale.dev/models/gpt-5-4-mini) | OpenAI | 63.70 | Observed 2026-11-21 · source benchlm:models |
| 42 | [Claude Sonnet 4.5 Thinking](https://modelscale.dev/models/claude-sonnet-4-5-thinking) | Anthropic | 63.67 | Observed 2026-11-21 · source benchlm:models |
| 43 | [GPT-5.4 nano](https://modelscale.dev/models/gpt-5-4-nano) | OpenAI | 51.50 | 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.
