I am trying to migrate from ElevenLabs, which does have a way to show history for requests. Not sure if you plan to have this as well. But in the short term so that when implemented I can come back, is the header x-request-id the identifier of the TTS request?
#x-request-id -- is this a unique ID for each TTS task?
limber egret
molten loom
Excited to help you migrate over @limber egret!
the x-request-id is typically used to keep a track of the HTTP connection so it won't be very helpful here! you should look into context-id which you can pass in and then we'll return it in every response corresponding to that request.
cc @fast kite @signal shoal
limber egret
Does this only work with Websocks @molten loom ?
molten loom
Also works on SSE!
limber egret
I am using the HTTP request. My goal is to know what the "job" is, similar to how Elevenlabls can use that request-id/history-id to pull information about what the job was. this was important because there were a number of instances that the audio output was bad and malformed and i needed it as a way to show them
this is the end point I am using. Are there docs on the headers and which I should use? https://docs.cartesia.ai/api-reference/tts/bytes
limber egret
I still don't see what the header value is:
@molten loom @fast kite
molten loom
We're working on adding support for getting TTS History on the /bytes endpoint very soon! cc @signal shoal @junior arch for helping you out with usage