← Latest products

Calling tools w/ Python improves LLM perf. vs MCP (77.1% on BrowseComp)

If agent's tools are exposed as functions/objects in a Python REPL (as opposed to JSON schemas) they perform better, I linked the explainer article we wrote, but if you want to jump straight in check out the docs!

Visit symbolica.aiexternal site

Reactions (1)

Similar products

by description similarity

  1. Forge

    github.com

    Guardrails take an 8B model from 53% to 99% on agentic tasks

    Hacker News687 pts

  2. MonkeyPatch

    github.com

    Cheap, fast and predictable LLM functions in Python

    Hacker News95 pts

  3. I built an MCP server using Cloudflare's code mode pattern

    github.com

    An implementation of the Codemode workflow detailed by Cloudflare.

    Hacker News87 pts

  4. PicoFlow

    news.ycombinator.com

    a tiny DSL-style Python library for LLM agent workflows

    Hacker News11 pts

  5. Llmswap

    pypi.org

    Python package to reduce LLM API costs by 50-90% with caching

    Hacker News12 pts