GenAiHub
Back
Task Planning & Knowledge Search
All 1,000 tasks

MCP Atlas

Real-world multi-step tool-use evaluation through the Model Context Protocol.

Published source(opens in a new tab)

Metric

Pass rate

Results shown

34

Publisher

Scale Labs leaderboard

Snapshot

Fetched 2026-10-23T00:21:15.119Z

34 of 34 models

34 of 34 models

Higher is better

Muse Spark 1.1 logo
Muse Spark 1.1
Fa
Fable 5.1
claude-opus-5 (xhigh) logo
claude-opus-5 (xhigh)
Qwen3.8-2.4T-A95B (xHigh) logo
Qwen3.8-2.4T-A95B (xHigh)
GLM 5.3 logo
GLM 5.3
Gemini 3.5 Flash (high) logo
Gemini 3.5 Flash (high)
Claude Fable 5 logo
Claude Fable 5
kimi-k3 (max) logo
kimi-k3 (max)
Muse Spark logo
Muse Spark
claude-opus-4-8 (max) logo
claude-opus-4-8 (max)
gpt-5.6 (sol) logo
gpt-5.6 (sol)
In
Inkling-small
claude-opus-4-7 (max) logo
claude-opus-4-7 (max)
gemini-3.1-pro-preview (high) logo
gemini-3.1-pro-preview (high)
glm-5p2 logo
glm-5p2
claude-opus-4-6 (max) logo
claude-opus-4-6 (max)
In
Inkling (xHigh)
glm-5p1 logo
glm-5p1
gpt-5.5 (xhigh) logo
gpt-5.5 (xhigh)
gpt-5.4 (xhigh) logo
gpt-5.4 (xhigh)
gemini-3-pro-preview logo
gemini-3-pro-preview
claude-opus-4-5 (high) logo
claude-opus-4-5 (high)
claude-sonnet-4-6 logo
claude-sonnet-4-6
gpt-5.2 (xhigh) logo
gpt-5.2 (xhigh)
kimi-k2p5 logo
kimi-k2p5
Ne
Nemotron 3 Ultra (thinking)
gemini-3-flash-preview logo
gemini-3-flash-preview
claude-sonnet-4-5 (thinking) logo
claude-sonnet-4-5 (thinking)
glm-4p7 logo
glm-4p7
gemini-3.1-flash-lite (high) logo
gemini-3.1-flash-lite (high)
gpt-5.4-mini (xhigh) logo
gpt-5.4-mini (xhigh)
gpt-5.1 (high) logo
gpt-5.1 (high)
o3-pro logo
o3-pro
claude-haiku-4-5 logo
claude-haiku-4-5
List bars visualize relative performance. Plot bars use published 0–100 scores, shrink to a 1.5rem minimum, then scroll horizontally.

Scores are published snapshots and are only comparable within the same benchmark, version, metric, and evaluation configuration.