MCP-Atlas
An evaluation suite built on the Model Context Protocol (MCP) ecosystem, designed to measure the overall performance of large models in real tool-calling scenarios. It covers capabilities including tool selection, parameter construction, multi-step invocation, and orchestration of external resources.
Timeline (last 90 days)
In the edge-cloud deployment scenario on MCP-Atlas, where the test involves three cloud models, SpecAgent closes 53% to 75% of the quality gap between the weak and strong models, while the number of tokens generated by the strong model is at most half of the original.
On the MCP-Atlas (tool calling) benchmark, Muse Spark Open scores 89.2 and GLM 5.3 scores 84.2.
Muse Spark Open leads existing open-source models in the MCP-Atlas evaluation.
MCP-Atlas is an evaluation collection built on the MCP ecosystem, used to measure a model's overall performance in real tool-calling scenarios.
All Facts (4)
In the edge-cloud deployment scenario on MCP-Atlas, where the test involves three cloud models, SpecAgent closes 53% to 75% of the quality gap between the weak and strong models, while the number of tokens generated by the strong model is at most half of the original.
50%UnverifiedOn the MCP-Atlas (tool calling) benchmark, Muse Spark Open scores 89.2 and GLM 5.3 scores 84.2.
50%UnverifiedMuse Spark Open leads existing open-source models in the MCP-Atlas evaluation.
50%UnverifiedMCP-Atlas is an evaluation collection built on the MCP ecosystem, used to measure a model's overall performance in real tool-calling scenarios.
50%