MCP Atlas leaderboard: who is #1

Name: MCP Atlas rankings
Creator: Scale AI
License: https://creativecommons.org/licenses/by/4.0/

Scale AI · Pass rate, higher is better · 34 models · published Sep 23, 2026

Scale AI

Top 25 of 34 models, best variant per model as published

#	Model	Score
1	Muse Spark 1.1Meta	88.1%
2	Claude Fable 5.1as “Fable 5.1”	87.2%
3	Claude Opus 5as “claude-opus-5 (xhigh)”	85.8%
4	Qwen3.8 2.4T A95Bas “Qwen3.8-2.4T-A95B (xHigh)”	84.5%
5	GLM-5.3as “GLM 5.3”	84.2%
6	Gemini 3.5 Flashas “Gemini 3.5 Flash (high)”	83.6%
7	Claude Fable 5Anthropic	83.3%
8	Kimi K3as “kimi-k3 (max)”	82.3%
9	Claude Opus 4.8as “claude-opus-4-8 (max)”	82.2%
9	Muse SparkMeta	82.2%
11	GPT-5.6as “gpt-5.6 (sol)”	81.8%
12	Inkling Smallas “Inkling-small”	79.2%
13	Claude Opus 4.7as “claude-opus-4-7 (max)”	79.1%
14	Gemini 3.1 Pro Previewas “gemini-3.1-pro-preview (high)”	78.2%
15	GLM-5p2as “glm-5p2”	77.8%
16	Claude Opus 4.6as “claude-opus-4-6 (max)”	76.8%
17	Inklingas “Inkling (xHigh)”	76.0%
18	GLM-5p1as “glm-5p1”	75.6%
19	GPT-5.5as “gpt-5.5 (xhigh)”	75.3%
20	GPT-5.4as “gpt-5.4 (xhigh)”	70.6%
21	Gemini 3 Proas “gemini-3-pro-preview”	70.3%
22	Claude Opus 4.5as “claude-opus-4-5 (high)”	69.8%
23	Claude Sonnet 4.6as “claude-sonnet-4-6”	69.5%
24	GPT-5.2as “gpt-5.2 (xhigh)”	67.6%
25	Kimi K2P5as “kimi-k2p5”	64.4%

About this board and its history

What it measures

Real tasks over 40+ MCP servers and 220 tools, three to six tool calls each.

Other agents and tools boards

Source

Scores as published by Scale AI on the capture date (2026-09-24). Source: labs.scale.com leaderboard page.

labs.scale.com/leaderboard/mcp_atlas

Sources: each leaderboard's own published results page, captured every day at 04:30 UTC. Scores exactly as each board published them; we run no evaluations. Logos via logo.dev; trademarks belong to their owners.

What it measures

Other agents and tools boards

Source

New #1s by email