You’ll need the gateway base URL (such as
https://your-gateway.example.com/v1) and an API key (it starts with
sk-olga-). Your operator creates these in Set up the
gateway.Send a request
Point your client at the gateway and use your gateway key. Everything else about your code stays the same.The gateway speaks the OpenAI Responses API at
/v1/responses and the
OpenAI Embeddings API at /v1/embeddings, plus the Anthropic Messages
API at /v1/messages. Use whichever your SDK targets. Streaming on the
text-generation endpoints behaves exactly as it does against the provider
directly: set stream=True (or stream: true) and read the events as usual.What you just got
On its way to the provider and back, your request was also:- Checked against your usage limits, and refused if it would exceed one.
- Screened by guardrails for PII and prompt injection, on supported content endpoints.
- Traced to your Openlayer project, with model, tokens, latency, and cost.

If something goes wrong
The most common errors are below. Errors covers every status code, the exact messages, and which ones are worth retrying.429 Too Many Requests
429 Too Many Requests
A usage limit on your key or group has been reached. Ask your operator to review your
usage limits.
422 Unprocessable Entity
422 Unprocessable Entity
A guardrail blocked the request or the response. The message is the guardrail’s reason, such as
“Detected high-risk PII entities: US_SSN”. See Guardrails.