Overview
Configure Meta's Llama models as the LLM provider for KamoCRM's SI assistant. Llama is open-weight β you can route through hosted providers (Together SI, Groq, Fireworks) or self-host on your own infrastructure for full data isolation.
Common use cases
Self-hosted SI for data sovereignty
Organizations requiring full control over SI inference can self-host Llama on their own GPUs with zero data egress.
Hosted Llama via Together, Groq, or Fireworks
Use hosted providers for Llama if self-hosting isn't practical.
Llama 3.1 405B for high-quality open SI
The largest Llama models rival GPT-4 and Claude Sonnet on many benchmarks while remaining open-weight.
Setup
- 1Choose hosting: self-host, Together SI, Groq, Fireworks, or another provider
- 2Obtain endpoint URL and API credentials
- 3Navigate to Settings β SI β Providers in KamoCRM
- 4Select Meta Llama as a provider
- 5Enter endpoint and credentials
- 6Test with a sample SI query
What data syncs
SI request/response routing
Token usage (via your chosen hosting provider)
Model availability
Questions about Meta (Llama)
Connect Meta (Llama) to KamoCRM on the free plan. No credit card required.
Start freeReady to connect Meta (Llama)?
Start your free KamoCRM account and connect in minutes.
