Benchmark
ARC-AGI Benchmark
What is ARC-AGI?
The Abstraction and Reasoning Corpus for Artificial General Intelligence (ARC-AGI) is a benchmark designed to test general intelligence and abstract reasoning…
- Released
- 2019-11-05
- Modality
- image
- Evaluates
- reasoning, spatial reasoning, vision
- Openness
- unknown
- Importer
- LLM Stats
- Reported scores
- 8
Source provenance
- Original evidence https://api.zeroeval.com/leaderboard/benchmarks/arc-agi?top_n=500
ARC-AGI results and reported scores
LLM Stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| GPT-5.2 | OpenAI | 0.862 | 2025-12-11 | evidence |
| GPT-5.2 Pro | OpenAI | 0.905 | 2025-12-11 | evidence |
| GPT-5.4 | OpenAI | 0.937 | 2026-03-05 | evidence |
| GPT-5.5 | OpenAI | 0.95 | 2026-04-23 | evidence |
| Inkling-Small | Thinking Machines Lab | 0.84 | 2026-07-30 | evidence |
| LongCat-Flash-Thinking | Meituan | 0.503 | 2025-09-22 | evidence |
| o3 | OpenAI | 0.88 | 2025-04-16 | evidence |
| Qwen3-235B-A22B-Instruct-2507 | Qwen | 0.418 | 2025-07-22 | evidence |
Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.