Hacker News · AI·· 12 h agoAI score58
Inference Gateway Desktop launches as a provider-agnostic AI agent client
Inference Gateway Desktop, an AI client for any model provider
AI summary
Inference Gateway Desktop is a native desktop app for AI agents that works with any model provider using your own API keys, covering OpenAI, Anthropic, Google, local Ollama models and any OpenAI-compatible endpoint. Agents run in a chosen project directory with file tools sandboxed to it, or on a Kubernetes workspace whose screen can be watched and taken over, with per-workspace switches for computer use, browser use, recording and text to speech. Releases are unsigned, so first launch needs an extra confirmation click, and on first run the app downloads the infer CLI and gateway into ~/.infer/.
Source:Hacker News · AI · github.com