Open the Model Hub in the console. You can filter by vendor, type, and tag. Click any model card to open its details:
Section
Description
Model name
The value for the model field. Use the copy button next to the name to avoid typos
Basic info
Capabilities, context length, input and output modalities
API endpoints
Protocols and paths the model supports, for example openai: /v1/chat/completions or gemini: /v1beta/models/{model}:generateContent
Group pricing
Discount and unit price (USD) for each group
API tab
Sample calls for this model
TIP
Check "API endpoints" before you call a model. Image and video models usually support only their own dedicated APIs and cannot be called through /v1/chat/completions.
A group determines which channels serve your requests and what price applies. The same model can cost different amounts in different groups.The following groups are currently available:
Group
Description
default (Official price)
All models available at official prices
Brand groups (such as openai, claude-aws, gemini, deepseek, qwen)
Discounted prices for that brand's models. See Group pricing on the model detail page for the exact discount
Choose a group type when you create a key:
Stability First / Price First: the system automatically picks among available groups following the auto-routing order shown on the model detail page. No manual choice needed.
Specify group: always use one group, useful when you want a fixed price.
Memberships and subscriptions
The Coding Plan page in the console offers subscription plans. After you subscribe, your account is upgraded to the plan's group for the plan period and gets lower prices; it returns to your original group when the plan expires. See the purchase page for each plan's credit, group, and duration.
Model credit: top up in Model Credit Top-up. All model calls are charged to this balance.
Search & Reader credit: purchase separately in MCP Products → MCP Plans & Top-up. It is not shared with your model balance.
Vouchers: view and use them in Voucher Management in the console.
Usage: Dashboard and Usage Logs in the console show spend by key, model, and time. You can also fetch it programmatically through the Balance and usage API.