Request headers - Altrum

None of these headers is required. Send them on any model request, in any format.

Header	Value	What it does
`x-request-id`	Any text	Your own reference for the request. Altrum uses it instead of making one, returns it, and shows it in the request log.
`x-alt-prompt-id`	A prompt id	Puts a saved prompt before your messages.
`x-alt-prompt-version`	A whole number, such as `2`	The prompt version to use. Without it, the latest version.
`x-alt-cache-key`	Any text	Keeps this request’s cached answers apart from others, for example per customer.
`x-alt-cache-ttl`	`30s`, `5m`, `1h`, or seconds	Keeps the answer for less time than your cache setting.
`x-alt-cache-threshold`	`0` to `1`	Asks for a closer match than your cache setting.
`x-alt-cache-type`	`direct` or `semantic`	`direct` matches word for word only, `semantic` by meaning only.
`x-alt-cache-no-store`	`true`	Serves from the cache when it can, but does not store this answer.
`x-alt-disable-content-logging`	`true`	Keeps this request’s prompt and reply out of the request log, even when your organization keeps content.
`x-alt-direct-key`	`true`	Sends your own provider key. See Authentication.

The cache headers can only narrow what your organization set: a request cannot keep an answer longer, or accept a looser match, than the cache settings allow. In the same way, x-alt-disable-content-logging can only turn content logging off.

Example

curl https://api.altrum.ai/v1/chat/completions \
  -H "Authorization: Bearer $ALTRUM_VIRTUAL_KEY" \
  -H "Content-Type: application/json" \
  -H "x-request-id: order-4512-summary" \
  -H "x-alt-prompt-id: support-reply" \
  -H "x-alt-cache-key: customer-881" \
  -d '{"model": "openai/gpt-4o-mini", "messages": [{"role": "user", "content": "Where is my order?"}]}'

In an SDK, set them as default headers on the client, or per request: see SDKs.