Create Custom Models
RecurseChat supports adding custom models in several ways:
- Local GGUF models - Use your own GGUF model files stored locally
- Hugging Face GGUF models - Download and use GGUF models from Hugging Face
- OpenAI-compatible endpoints - Connect to services like Ollama, Azure OpenAI, Groq, and xAI Grok
Each option provides different benefits:
- Local GGUF models give you full control and privacy by running models on your own machine
- Hugging Face integration makes it easy to access a wide variety of open source models
- OpenAI-compatible endpoints let you use cloud services with enterprise features and the latest models
Local GGUF Models
For detailed instructions on adding local GGUF models, see the Local GGUF model guide.
Hugging Face GGUF Models
For detailed instructions on downloading and using GGUF models from Hugging Face, see the Hugging Face GGUF model guide.
OpenAI-Compatible Endpoints
For detailed instructions on adding OpenAI-compatible endpoints, including Ollama, Azure OpenAI, Groq and xAI Grok, see the OpenAI compatible endpoint model guide.