The LiveKit Agents framework supports a wide range of large language models (LLMs) that can generate code for LiveKit APIs and features. Recommended models for code generation and integration with LiveKit include: • DeepSeek Coder models: Specifically designed for code tasks. • OpenAI GPT-4 series: Suitable for complex code and API integration. • Claude models (Anthropic): Good for detailed reasoning with API documentation. • Meta Llama CodeLlama variants: Trained for code generation, available through providers like Together AI. These LLMs are available via providers such as OpenAI, Anthropic, Google Gemini, Cerebras, Groq, Fireworks, and Together AI, with unified support in the LiveKit Agents framework. For self-hosting, Ollama also supports Llama and DeepSeek models. All supported models integrate seamlessly for building applications with LiveKit APIs, and you can switch between them as needed for optimal results. For realtime and voice-based coding scenarios, models like OpenAI Realtime API, Gemini Live, and Amazon Nova Sonic are also supported. Let me know if you want detailed setup instructions for a specific model or provider. Sources: Large language model (LLM) integrations | LiveKit Docs | livekit-plugins/livekit-plugins-openai/livekit/plugins/openai/llm.py | Llama and LiveKit | LiveKit Docs | LLM Integrations | livekit/livekit_composite | DeepWiki