Embeddings

POST /v1/embeddings — vector representations of text, billed and routed the same way as chat completions.

POST ${GATEWAY_URL}/v1/embeddings returns vector representations of your input text, using the same request and response shape as OpenAI's embeddings API.

POST https://gateway.relixr.com/v1/embeddings
Authorization: Bearer nf_live_...

{
"model": "text-embedding-3-small",
"input": "Relixr routes inference for agent fleets."
}
Tip:

Embedding calls are usually cheap and high-volume — a good candidate for a dedicated key with a tight daily spend cap, so a bug in a batch job can't run away with your budget.

Next steps
Chat completions
POST /v1/chat/completions
Billing & credits
How embedding costs are reserved and settled.
← Chat completionsAuthentication →