Off Grid AI
Run local models on your phone or computer for chat, document analysis, image generation and speech transcription. Connect to your own model server.
Off Grid AI runs language models on iOS, Android, macOS, Windows and Linux. You can chat, analyze documents and generate images on your own hardware. The mobile app uses the MIT license, while the desktop app uses AGPL.
The app supports GGUF models such as Llama, Qwen 3, Gemma 3, Phi-4 and Mistral. Inference uses the CPU or a supported GPU, including Metal on Apple Silicon and Adreno GPUs on Android. If a model is too large for your phone, you can connect to Ollama, LM Studio, LocalAI or another OpenAI-compatible server on your local network. The model then runs on that server.
Stable Diffusion generates images on-device. Vision models can answer questions about photos, and Whisper transcribes spoken input locally. You can attach PDFs, CSV files and code to a conversation, or search saved documents through a local knowledge base.
Downloaded local models work offline. Web search and connected services need network access, and using a remote model sends requests to the server you select.
