Getting Started
Quickstart
Sythoria is a lightweight, privacy-first AI chat interface that connects to any OpenAI-compatible API. No accounts, no tracking — just pick a provider, add your key, and start chatting.
Installation
- Download the latest release from GitHub
- Install and launch Sythoria
- Click Start on the welcome screen
- Open Settings (gear icon in the sidebar) and configure your first model
- You're ready to chat
Add your first model
When you open Settings for the first time, a default OpenAI model is pre-filled. You can keep it, switch providers, or add more. For each model you need a provider preset and an API key — the rest fills in automatically.
| Provider | Setup guide | Where to get your key | Key? |
|---|---|---|---|
| OpenAI | /docs/providers/openai | platform.openai.com/api-keys | Yes |
| Anthropic | /docs/providers/anthropic | console.anthropic.com | Yes |
| Google Gemini | /docs/providers/gemini | aistudio.google.com/apikey | Yes |
| Ollama | /docs/providers/ollama | None — just run ollama server | No |
| OpenRouter | /docs/providers/openrouter | openrouter.ai/keys | Yes |
| NVIDIA NIM | /docs/providers/nvidia-nim | build.nvidia.com | Yes |
Your API key is stored only in the app's local storage — it never touches Sythoria's servers. See Privacy & Security for the full breakdown.
Start chatting
- Pick a model from the dropdown in the chat input bar
- Type a message and press
Enter - Responses stream in real-time, token by token
That's it. You're running Sythoria.
What's next?
- Streaming Responses — how real-time streaming works under the hood
- Multi-Provider Chat — switch between providers mid-conversation
- Configuration — temperature, system prompts, keyboard shortcuts, and more
- Privacy & Security — how your data stays safe
