Loading deployments…
Deployments
Live status · Read-only
OpenCode configuration
Select models for Mac-Local. Only running models can answer requests. Sign in to manage deployments.
Context defaults to the model configuration maximum. Output uses its declared limit, or Context minus one token when no separate limit is declared. This is a theoretical ceiling: input shares the same budget, and large contexts need more RAM.
Estimated prices from the administrator are shown when available. These are internal estimates, not measured operating costs or official pricing.
Limits the initial connection only. Long responses have no fixed time limit; the gateway checks model health while waiting. Copy the updated configuration and restart OpenCode to apply.
The key is included only in the generated configuration. If blank, replace YOUR_API_KEY_HERE before connecting. Do not share the generated file with other users.
For an existing configuration, merge the generated provider entry into its provider section. OpenCode documentation