Use any AI model with YUNA: local or via API
YUNA is a voice AI companion with a Live2D face. The "brain" is up to you: run a model fully on your PC with Ollama, or connect your own API key.
Run locally with Ollama
Any Ollama model works out of the box. If it runs on Ollama, it runs in YUNA, with no configuration. Whatever you have pulled (Qwen, DeepSeek-R1, Llama, Gemma, Mistral and others) appears automatically. Nothing leaves your machine and no API key is needed.
Connect a cloud provider with your own API key
- OpenAI: use any GPT model your key can access.
- Anthropic (Claude): use any Claude model your key can access.
- Google Gemini: use any Gemini model your key can access.
- DeepSeek: reasoning and chat models.
- Groq: fast hosted open models such as Llama and Mixtral.
Enter the key in settings, pick the model you like and switch whenever you want.
Local or cloud: which should you pick?
| Local (Ollama) | Cloud (API key) | |
|---|---|---|
| Privacy | Everything stays on your PC | Messages go to the provider |
| Internet | Only for the first model download | Required |
| Cost | Free after you buy YUNA | Provider usage fees |
| Speed | Depends on your hardware | Fast on any PC |
See the system requirements or the FAQ.