# MCP Atlas — AI model rankings

Can the AI chain together many tools and steps to complete one bigger task, rather than doing just a single thing? Higher is better.

11 tracked models have a published MCP Atlas score. Higher is better. Scores come from published lab reports and benchmark sources; recorded sources appear beside the scores.

## Ranking

| Rank | Model | Developer | Score | Source | Released |
| --- | --- | --- | --- | --- | --- |
| 1 | Muse Spark 1.1 | Meta | 88.1% | Lab | Jul 9 2026 |
| 2 | Kimi K3 | Moonshot AI | 84.2% | Lab | Jul 16 2026 |
| 3 | Gemini 3.5 Flash | Google | 83.6% | Lab | May 19 2026 |
| 4 | Muse Spark | Meta | 82.2% | Lab | Apr 8 2026 |
| 5 | Claude Opus 4.7 | Anthropic | 79.1% | Lab | Apr 16 2026 |
| 6 | Gemini 3.1 Pro | Google | 78.2% | Lab | Feb 19 2026 |
| 7 | Qwen3.7-Max | Qwen | 76.4% | Lab | May 20 2026 |
| 8 | Muse Glimmer | Meta | 75.5% | Lab | Aug 10 2026 |
| 9 | GPT-5.5 | OpenAI | 75.3% | Lab | Apr 23 2026 |
| 10 | Claude Sonnet 4.6 | Anthropic | 69.5% | Lab | Feb 17 2026 |
| 11 | Gemini 3.0 Flash | Google | 62% | Lab | Dec 17 2025 |


---

Canonical page: https://aireleasetracker.com/benchmark/mcp-atlas
Full dataset: https://aireleasetracker.com/llms-full.txt · JSON: https://aireleasetracker.com/models.json
Source: AI Release Tracker (https://aireleasetracker.com). Most benchmark scores come from lab launch material; gathered results identify the leaderboard that published them.