Rankings
Snapshot 22 November 2026 · Scores copied from each benchmark's own leaderboard · How the CTO Index works
Score and price
| # | Model | Provider | ARC-AGI-3 | Price $/M |
|---|---|---|---|---|
| 1 | Claude Fable 5.2 | Anthropic | 81.2% | $8 / $40 |
| 2 | GPT-6.2 Astra | OpenAI | 70.3% | $8 / $40 |
| 3 | Gemini 4 Carbon | 65.1% | $2 / $10 | |
| 4 | GPT-6 Astra | OpenAI | 62.7% | $10 / $50 |
| 5 | GPT-6.2 Sol | OpenAI | 54.9% | $2 / $10 |
| 6 | GPT-6.1 Sol | OpenAI | 52.7% | $2 / $10 |
| 7 | Gemini 4 Argon | 49.8% | $2 / $10 | |
| 8 | Claude Opus 5 | Anthropic | 30.2% | $5 / $25 |
| 9 | Claude Sonnet 5.5 | Anthropic | 24.3% | $2 / $10 |
| 10 | Gemini 3.8 Flash | 10.4% | $0.75 / $3.75 | |
| 11 | GPT-5.6 Sol | OpenAI | 7.8% | $4 / $20 |
| 12 | GPT-6 Sol | OpenAI | 4.6% | $2 / $10 |
| 13 | Grok 4.6 | xAI | 2.1% | $1.60 / $4.80 |
| 14 | Grok 4.7 | xAI | 1.8% | $2 / $6 |
| 15 | Claude Opus 4.8 | Anthropic | 1.5% | $5 / $25 |
| 16 | Claude Haiku 5.5 | Anthropic | 0.5% | $0.10 / $0.50 |
| 17 | Inkling 2 Small | Thinking Machines | 0.3% | $0.30 / $1 |
| 18 | DeepSeek V4.1 Flash | DeepSeek | 0.2% | $0.15 / $0.60 |
| 19 | GPT-6 Luna | OpenAI | 0.2% | $0.10 / $0.50 |
Other benchmarks
Frequently asked questions
What does ARC-AGI-3 measure?
Interactive game-like environments the model has to explore, learn the rules of and solve, scored on the standard (no provider adapter) track.
Which model has the highest ARC-AGI-3 score?
Claude Fable 5.2 from Anthropic leads ARC-AGI-3 at 81.2%, ahead of GPT-6.2 Astra (70.3%).
Why does ARC-AGI-3 matter to engineering leaders?
It tests learning on the job inside an environment, which is what long-running autonomous agents have to do.
What are the limitations of ARC-AGI-3?
Only a handful of frontier models have been run, and costs run into thousands of dollars per evaluation, so we show it but leave it out of the CTO Index.