MCP Atlas public (Gemini 3.1 Pro report)
Agents · 2026-02-19
Historical public-set MCP-Atlas results that Google's Gemini 3.1 Pro methodology attributes to Turing. The comparison table includes original-protocol Opus 4.6 59.5 and GPT-5.2 60.6. Judge, retry handling and execution budget are not specified for this source; retain this public cohort separately from later Scale runs.
Top models (higher is better)
| Model | Score |
|---|---|
| Gemini 3.1 Pro Preview | 69.2 |
| Sonnet 4.6 | 61.3 |
| GPT-5.2 | 60.6 |
| Opus 4.6 | 59.5 |
| Gemini 3 Pro Preview | 54.1 |