← Back to Documentation
Models & Providers
Local Model Integration
Connect llama.cpp, OpenAI-compatible gateways, or other production-ready internal model endpoints.
Key Steps
- Expose the local model endpoint through a controlled server URL.
- Add it as a private provider and map its model IDs to clear aliases.
- Measure latency, throughput, and hardware saturation before production use.
Article Scope
This guide focuses on the operational steps needed to run Orchestris with team access, provider control, and predictable usage.
Visual Reference
Endpoint requirements
- Use a stable base URL reachable from Orchestris Server.
- Protect the endpoint with network controls and credentials where possible.
- Confirm the endpoint follows the expected OpenAI-compatible request and response shape.
Model registration
- Add the endpoint as a private provider in the catalog.
- Create user-facing aliases that describe the model purpose.
- Limit access to teams that understand the model quality and data constraints.
Capacity planning
- Test concurrency with representative prompts.
- Monitor queue time, latency, and failed requests.
- Document fallback behavior if the local model host is unavailable.
Related Topics
Models & Providers
Using AI Models
Help users choose the right models without exposing provider complexity.
Models & Providers
OpenAI Integration
Connect OpenAI models and expose approved options through the provider catalog.
Models & Providers
Anthropic Integration
Connect Anthropic models for teams that need strong long-context and reasoning workflows.