Downloads
Pick your OS.
Native desktop builds for macOS, Windows, and Linux. Code-signed and auto-updating.
Beta builds. Auto-update is on by default. Code signing for macOS and Windows.
System requirements
Modest. Local-first.
Entity runs on the hardware you already have. Network use is opt-in and only points at your own model endpoint.
Disk
~600 MB
Memory
4 GB minimum, 8 GB recommended
Network
Optional — only used to reach your model endpoint
Models
Any OpenAI-compatible endpoint, Ollama, vLLM, or BYO API key
After you install
Plug in a model. Start.
Three ways to point Entity at the model you want to use. Pick whichever matches the work.
Use a local model
Point Entity at your Ollama or vLLM server. No keys, no cloud.
Use a premium API
Paste an API key for OpenAI, Anthropic, Gemini, or any compatible endpoint.
Use your fine-tune
Any OpenAI-compatible endpoint works. Self-hosted, hosted, your call.