dsh-model-reasoning

When using DSH with custom model providers, you must manually fill in reasoning effort, context window, max tokens, and image modality support for each model one by one — tedious and easy to miss. dsh-model-reasoning eliminates this burden by pulling authoritative data from models.dev and auto-populating all these parameters for every non-official provider, so you no longer need to look up docs and type configs manually.

The plugin uses a local-cache-first strategy: at startup it fills parameters immediately from a bundled cache instead of waiting on the network, then asynchronously fetches the latest data to keep the cache fresh. It also listens for model configuration changes and re-fills automatically, while preserving any values you have already set so your customizations are never overwritten. A visual settings card in the web panel lets you toggle auto-fill and data-sync with a single click and apply changes instantly.

Ideal for developers working with multiple custom providers who want to skip repetitive parameter setup. Configuration lives in a dedicated namespace, so plugin upgrades migrate old settings automatically and rolling back to a previous version is seamless — no manual migration required at any point.