Available Models
Available Models lists the models you can currently call. Use it to find model IDs, compare context windows and prices, and see whether a model has been running normally. Open the page from the console's left navigation.
The list scope is configured by your administrator; different members may see different models.
Finding the Model You Need
Models are shown as cards, with search, filter, and sort controls at the top:
- Search: Use Search by model name, ID, or vendor; the list filters as you type.
- Vendors: Show only one vendor's models; choose "All" to restore the full list.
- Sort: Pick a sort key in Sort by (for example, Sort by: Name), then use the button on the right to toggle ascending or descending.
With many models, narrow by vendor first, then search for a specific one.
Reading a Model Card
Each card is one model.
| Card item | Description |
|---|---|
| Display name | The readable model name, shown as the card title |
| Model ID | The short name under the display name—the value for the request's model field; click the icon on the right to copy |
| Description | The model's capabilities and use cases |
| Context | The maximum length the model can handle in a single request |
| Input Price / 1M Tokens | Price per million input tokens |
| Output Price / 1M Tokens | Price per million output tokens |
| Cache Write / Read Price | The model's prompt cache write and read prices |
| Health · Last 24 Hours | 24 small blocks, one per hour |
Per-request models show a price per call in the input price slot, with — for output. Any -- means the item is not configured.
Health · Last 24 Hours runs left to right in time order, ending at the current hour. Hover a block to see the time slot and success rate:
| Color | Meaning |
|---|---|
| Green | Calls in that hour were mostly normal |
| Yellow | A noticeable share of failures in that hour |
| Red | Many failures in that hour |
| Gray | No calls or no monitoring data for that hour |
The status reflects the model's recent overall activity, not your personal requests. A gray block does not mean the model is unavailable—it usually just means nobody called it during that hour.
Calling a Model
- Find the target model and click the copy icon next to its model ID.
- Put the copied ID into the request's
modelfield and call with your API key; see Quickstart for the request format. - Confirm the result, token usage, and cost in Request Logs.
Always use the model ID, not the display name on the card.