Skip to main content
Embeddings use the OpenAI-style /v1/embeddings shape.

Request

Response

Metering

Embedding calls are metered from input usage. The adapter normalizes provider output into an embeddings array plus usage.