Models
The left sidebar on Manage Models lists providers and their models. Each row shows context window and input price. Badges mark Default (seeded) vs Custom (added by you). Click a model to open the editor on the right, or use + on a provider to add one.
Edit or add a model
Section titled “Edit or add a model”| Field | Notes |
|---|---|
| Model ID | Identifier your SDK sends (e.g. gpt-4o-2024-08-06). Normalized to lowercase; only editable when adding. |
| Display Name | Human-readable name in dropdowns. |
| Model Type | Chat, Embeddings, Images, or Audio - controls how export JSON is shaped. |
| Context Window | Max context length in tokens. |
| Input Price (per 1M tokens) / Output Price (per 1M tokens) | Used for AI cost tracking and recalculation. |
| Capabilities (comma separated) | Optional list (e.g. function-calling, vision, streaming). |
Click Save Model to persist. For an existing model, Delete opens a confirmation titled Delete Model.
Empty state copy when nothing is selected: Select a model to view details.
- Model ID should match
gen_ai.request.modelon your spans so Configuration and ingest-time pricing can resolve it. - After changing prices, use Auto Pricing or the Cost tile refresh icon on a trace for historical spans; the SDK Pricing URL applies to future tracing.
List, create/update, and delete models