Choosing models
Choose a model for your v0 chats, add models from AI Gateway, and customize supported model settings.
Use the model picker in the chat composer to choose the model that responds to your prompts. You can use v0 models or add custom models from the available AI Gateway catalog.
The updated model picker is available to Enterprise teams. Custom models require a team owner's acceptance before members can use them. See Enterprise model access.
Choose a model
- Open a chat in the workspace you want to use.
- Click the current model in the composer toolbar.
- Select an available model, then send your prompt.
Available models depend on your plan and workspace settings.
Add custom models
- Open the model picker and select Add models.
- Browse the catalog or search by model name or provider.
- Select a model to add it to your picker and open its customization panel.
- Review the available settings, then use Back to models to return to the picker and select the model.
Adding a model to your picker does not enable it for everyone in your workspace.
Customize a model
Open a custom model's customization panel from the picker to review its pricing and supported options:
- Thinking: Choose a reasoning effort when the model supports it. Available levels vary by model.
- Fast Mode: Enable faster responses on models that offer this option. Review the displayed pricing when changing modes.
- Remove model: Remove an added model from your picker.
Model prices vary. Review input and output token prices in the picker and see Pricing for how credits are used.
Locked models
On Enterprise teams, custom models are locked until a team owner enables Allow custom models in Workspace settings. Members can browse locked models but cannot use them. v0 models remain available.
If a custom model is locked, ask a team owner to review custom model access. Workspaces with isolated tenant restrictions remain limited to their supported v0 models.
To add AI functionality to an app you are building, see AI models.