Response Headers
ai& exposes request metadata through HTTP headers rather than modifying the OpenAI- or Anthropic-shaped response body. This preserves wire-format compatibility with the official SDKs.
Identity & tracing
Section titled “Identity & tracing”| Header | Description |
|---|---|
X-Request-ID | Unique ID for this request. Include in support tickets. |
X-AiAnd-Org-ID | Organization the request was billed to. |
X-AiAnd-Model | Model that actually served the request (after any aliasing). |
Billing
Section titled “Billing”| Header | Description |
|---|---|
X-AiAnd-Cost-USD | Cost of this request in USD. Set on every billable response. |
X-AiAnd-Input-Tokens | Input tokens counted toward billing. |
X-AiAnd-Output-Tokens | Output tokens counted toward billing. |
Rate limiting
Section titled “Rate limiting”See Rate Limits for the full set of X-RateLimit-* headers.
Streaming
Section titled “Streaming”Streaming responses carry the identity headers immediately but emit cost/tokens via a trailer event after the model’s terminal message. See Streaming Events.