我将Google TimesFM-3包装为MCP服务器

2026年9月2日1 次浏览来源:Dev.to阅读原文

正文保留英文原文(机翻易破坏代码与排版),标题/摘要已提供中文

Google released TimesFM-3 on 31 August

2026.

I shipped an MCP server for it the next day so Claude Desktop, Claude Code, and Cursor can call the model as a tool.

Repo: https://github.com/thenameisdevair/timesfm3-mcp What the agent actually gets The tool is .

You pass a chronological list of numbers.

You get: — median / point prediction — the nine official TimesFM-3 heads, through a license note on every response TimesFM-3 already produced those quantile heads.

Most thin wrappers throw them away.

This one does not.

The license is the product constraint Piece License This MCP wrapper Apache-2.0 TimesFM-3 weights () non-commercial, non-production You can research, evaluate, and run agent experiments.

You cannot put this checkpoint behind a paid API, a customer deliverable, or a live demand planner.

Google’s commercial path is BigQuery / AlloyDB .

If a README hides that, it is not a serious integration.

Install Weights are gated on Hugging Face.

Accept the model terms first.

Point Claude or Cursor at local/server.py using the venv Python.

Config snippets are in the README.

What this release is not No multivariate targets yet.

No past/future covariates.

No commercial backend.

Those are next.

This cut exists so “TimesFM-3 MCP” has a public, honest surface while the launch is still warm.

If the tool shape is wrong for your agent, feel free to open an issue on the repo.

分享