Skip to content

Can I use local models with coding agents like Claude Code, Codex or OpenCode?

Often, yes. Many coding agents let you set a custom API base URL, and local servers such as exo expose OpenAI- and Anthropic-compatible endpoints. Point the agent at your local endpoint and choose a large open model that is strong at tool calling, such as a recent GLM, Kimi, DeepSeek or Qwen coding model.

OpenCode and the Codex CLI support custom providers through their config files. Claude Code reads ANTHROPIC_BASE_URL, which Anthropic documents for LLM gateways; it is designed for Claude models, so results with other models are not supported by Anthropic.

Coding agents send long prompts with many files. Prompt processing speed (prefill) and context length matter as much as generation speed. Test with a real task from your codebase.

Smaller models often fail at multi-step tool use. For agentic coding, prefer the largest model your hardware runs at a usable speed.

Sources

  1. exo on GitHub (README)
  2. Claude Code docs: LLM gateway configuration
  3. OpenCode docs: providers

Updated

Planning local AI for your team?

Tell us about your models, users and hardware. Exo Labs helps teams size, build and run private AI clusters.

Talk to Exo Labs →Download exoCheck your hardware on local.ai ↗