Blog
Product updates, guides, and engineering notes from the team.
Why Your AI Agents Need a Local LLM Gateway
Running Codex, Claude Code and Gemini CLI side by side? A local LLM gateway gives every agent one endpoint with model routing, failover, protocol conversion, observability and privacy. Here is how it works and how to set one up.
Run Codex CLI with Any Model: OpenRouter, DeepSeek, Ollama
Codex speaks the OpenAI Responses API. Learn how to run it on OpenRouter, DeepSeek, Ollama or any OpenAI-compatible model with a local gateway that converts Responses to Chat Completions, and what gets written to ~/.codex/config.toml.
How to Use DeepSeek (or Any Model) in Claude Code
Run Claude Code on DeepSeek, OpenRouter, Ollama or any OpenAI-compatible model through a local gateway. One-click setup, the exact settings.json changes, protocol conversion and routing explained.
What Is Jev? TypeSafe AI's Decision Model Explained
Learn how TypeSafe AI's Jev model returns typed choices, probabilities and scores through the Jev API, and where it fits in AI agent workflows.