Overview
Llama is a family of open source LLMs developed by Meta AI. These models are widely deployed by a large number of cloud inference providers, and also support your own local or private deployments.
Providers
LiveKit Agents has out-of-the-box support for all of these Llama inference providers:
Cerebras
Fast text-only inference for the latest Llama models.
Groq
Fast inference for the latest Llama models.
Fireworks
Inference, fine-tuning, and multimodal support for the latest Llama models.
Perplexity
The Sonar family of models are based on Llama 3.1, fine-tuned for search.
Telnyx
Hosted inference for the latest Llama models.
Together AI
Inference, fine-tuning, and multimodal support for the latest Llama models.
Ollama
Run Llama and other open source models locally.