Tools ·
Ollama Now Lets Codex Desktop Run Local AI Models
Tools ·
Ollama Now Lets Codex Desktop Run Local AI Models
AI summary · Generated from this article
Ollama now officially integrates with OpenAI’s Codex desktop app, letting developers use local or Ollama Cloud models while retaining Codex’s coding workflow. The integration requires Ollama v0.17.7 or later and Codex App 26.311.2136 or later, and launches with “ollama launch codex-app.” It uses Ollama’s OpenAI-compatible endpoint at localhost:11434, with Qwen3.5 as the documented example. Codex retains browser annotations, diff reviews, repository tools, command execution, and isolated Git worktrees; broader ChatGPT plugin compatibility is not confirmed.