Model Hub & Configuration
Download local models and configure cloud providers in one place.
The Model Hub is the central place to manage which models CrestBid can use — both local models on your GPU and, on paid tiers, cloud providers.
Local models
- 1Open Settings › Model Hub.
- 2Browse available local models and download the ones you want.
- 3Set a default model for new projects.
Pick a model size that fits your VRAM — see System Requirements.
Cloud providers (paid tiers)
Cloud and remote providers are an explicit, operator-controlled opt-in: they are configured server-side in the deployment's environment file with a credential you supply, and are off by default. There is no silent fallback to cloud — if a cloud provider isn't configured, nothing ever leaves your machine.
- Point generation and chat at a customer-supplied API endpoint with your own key, or
- Point the orchestrator at an OpenAI-compatible inference host elsewhere on your own network (remote model mode).
Whenever a cloud model is active, prompts leave your machine for that provider. Keep local models selected for sensitive work — see
Multi-Model Support.