- Completely works offline (except for LLMs that require external API)
- OpenAI, Anthropic, Google AI, Groq, Ollama etc
- Multi-modal
- Text-to-speech powered by OpenAI and ElevenLabs
- Text files: all models
- PDF: natively supported in Claude 3.5 Sonnet and Gemini, in other models we extract the text and inject as text prompt
- Audio: natively in Gemini and GPT-4o Audio
- Images: GPT-4o family, Gemini, Claude, Groq Lllama
For macOS and Windows. Mobile app soon(ish)