LLMs and models
Nexus supports the model catalog your policy allows. The lead story is Private AI models — self-hosted or partner-hosted Private AI as the organizationally aware default.
Catalog at a glance
| Category | What it is |
|---|---|
| Custom LLMs (Private AI) | OpenAI-compatible or Ollama endpoints you or a partner operate — the primary path for regulated / confidential workloads |
| Public LLMs | Cloud providers (OpenAI, Anthropic, Perplexity, …) — optional, often ranked as fallback |
Ranking and path
- Drag to rank providers; the first enabled provider is the default
- Prefer Diode connection modes (Perimeter Discover or manual bind) so the model hop stays on your private path
- Combine with Tools and MCP so Private AI is organization-aware, not isolated
Setup: Configure an LLM.