Trying to reach tts model and getting 404

Question

Hello all,

I am working on AzureOpenAI, i have a deployment on sweden, and have (amongst others) 2 models deployed: whisper and tts.

When i build a client with

client=AzureOpenAIclient(azure_endpoint,api_key,api_version, azuredeployment = Whisper_deplo_name)

and call it with client.audio.transcriptions.create(model=whisper-1)

No issues it works perfectly i received text.

But if i use

client=AzureOpenAIclient(azure_endpoint,api_key,api_version, azuredeployment = **tts_**deplo_name)

and call it with

client.audio.speech.create(model=tts-1,...)

I get a 404, no resource found error, i am flabbergasted by this error. Does anyone has any idea, or encountered the same type of issues?

Best regards

Answer 1

Ok obviously it's after writing the question that i find the answer even though i searched for hours before.
I will write it here if it can help someone:

MIND THE API VERSION!! it my vary on the model you use: here is the website to stay updated:

https://learn.microsoft.com/en-us/azure/ai-services/openai/api-version-deprecation

Answer 2

Solution proposée et validée par Louis GOMEZ

ATTENTION À LA VERSION API !! cela peut varier selon le modèle que vous utilisez : voici le site Web pour rester à jour :

https://learn.microsoft.com/en-us/azure/ai-services/openai/api-version-deprecation