Docs

What each part of goinfer does and how to use it. These pages are the repo's own documents, shown here so they are easy to find.

Start here: Use it from Go covers installing the module, the smallest working program, and output shaped like a Go struct. If you would rather run a server and call it over HTTP, start with the server reference instead.

Use it from Go

• Use it from Go

Serving

• HTTP server — OpenAI, Anthropic, and the rest of the surface
• serve flag reference
• API stability tiers — what v1.0 semver-binds

Connect a tool

• Claude Code → goinfer
• opencode → goinfer
• jevx and TypeSafe's SDKs → goinfer

Models and memory

• Quantization: what goinfer stands behind, and why
• Prequantized weight bundles (`.giw`)

The book

An inference primer for Go engineers: eleven chapters on how a language model runs, each ending in a number measured in this repo.