BlogLLM Benchmark Scores as Published, Side by Side
ARC-AGI-1: Results as Published
Every result the catalog holds for ARC-AGI-1: 13 models from 3 developers, read from 5 pages. Each number is as printed by the company named beside it; nothing is ranked or rescored. Models are listed by developer, newest first.
Area General Knowledge and ReasoningMaintainer ARC Prize FoundationIntroduced 2019Data version 2026103001, as of 2026-10-30
ARC-AGI-1 is a set of colored grid puzzles where the model sees a few input and output examples and must work out the rule for a new input. Its creator designed it to measure how efficiently a system learns a new skill from very few examples, rather than how much it has memorised. It was the first of the ARC-AGI series and was not beaten by AI systems for about five years after it appeared.
ARC-AGI-1
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPT-6 AstraOpenAI, released 2026-09-03 | 98.5% | OpenAI, GPT-6 Astrachecked 2026-10-14 | - |
| GPT-5.6 SolOpenAI, released 2026-07-09 | 97.5% | OpenAI, GPT-6 Astrachecked 2026-10-14; printed on 1 more page | - |
| GPT-5.5OpenAI, released 2026-04-23 | 95% | OpenAI, Introducing GPT-5.5checked 2026-10-15 | Effort: xhigh |
| GPT-5.4OpenAI, released 2026-03-05 | 93.7% | OpenAI, Introducing GPT-5.4checked 2026-10-15; printed on 1 more page | Effort: xhigh |
| GPT-5.4 ProOpenAI, released 2026-03-05 | 94.5% | OpenAI, Introducing GPT-5.4checked 2026-10-15; printed on 1 more page | Effort: xhigh |
| GPT-5.2OpenAI, released 2025-12-11 | 86.2% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 1 more page | Effort: xhigh |
| GPT-5.2 ProOpenAI, released 2025-12-11 | 90.5% | OpenAI, Introducing GPT-5.2checked 2026-10-15; printed on 1 more page | Effort: xhigh |
| GPT-5.1OpenAI, released 2025-11-12 | 72.8% | OpenAI, Introducing GPT-5.2checked 2026-10-15 | Effort: high |
| Claude Fable 5.1Anthropic, released 2026-09-01 | 97.5% | Anthropic, System Card: Claude Fable 5.1 & Claude Mythos 5.1checked 2026-10-25; printed on 1 more page | Effort: max |
| Claude Opus 5Anthropic, released 2026-07-24 | 97.5% | Anthropic, System Card: Claude Fable 5.1 & Claude Mythos 5.1checked 2026-10-25; printed on 1 more page | Effort: max |
| Claude Fable 5Anthropic, released 2026-06-09 | 98.5% | Anthropic, System Card: Claude Fable 5.1 & Claude Mythos 5.1checked 2026-10-25; printed on 1 more page | Effort: max |
| Claude Opus 4.7Anthropic, released 2026-04-16 | 93.5% | OpenAI, Introducing GPT-5.5checked 2026-10-15 | - |
| Gemini 3.1 Pro PreviewGoogle, released 2026-02-19 | 98% | OpenAI, Introducing GPT-5.5checked 2026-10-15 | - |
Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.