apifable and sufetch
These two tools are competitors, both providing an MCP server to facilitate AI-driven exploration of OpenAPI specifications and the generation of type-safe API clients.
About apifable
ycs77/apifable
MCP server that helps AI agents explore OpenAPI specs, search endpoints, and generate TypeScript types.
Implements MCP tools for structured API exploration—`list_endpoints_by_tag`, `search_endpoints`, and `get_endpoint`—with fuzzy search fallback and automatic schema reference resolution. Supports both local OpenAPI files and remote URLs with header-based authentication, using a config-driven approach where specs can be fetched and refreshed independently from git-tracked configuration. Generates self-contained TypeScript types from schemas or endpoint request/response structures for direct use in frontend integration code.
About sufetch
productdevbook/sufetch
Type-safe OpenAPI clients with MCP server for AI-driven API exploration
Generates fully-typed TypeScript clients from OpenAPI specs using apiful, then exposes them through an MCP server with tools like `search_endpoints` and `generate_code_example` for AI-assisted API exploration. Includes pre-built clients for DigitalOcean, Hetzner Cloud, and Ory services, with type helpers to extract request bodies, responses, and parameters for strict compile-time safety. Integrates with Claude Desktop and Claude Code CLI via stdio transport for seamless AI-driven code generation.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work