# Embeddings

> POST /v1/embeddings exists for compatibility, but no embedding model is available yet.


`POST /v1/embeddings` accepts the OpenAI body so that tools which probe it get a clear answer. Today every call returns:

```json
{"error": {"type": "not_found_error", "code": "model_not_found", "message": "no embedding model is available"}}
```

with status `404`. No credits are charged. When an embedding model is added, it will appear in `GET /v1/models` and in the [changelog](/en/references/changelog).

## Next

- [Models](/en/getting-started/models)