I’m excited to use the new GPT-4 Turbo model with “better writing”, but we’re migrating to Microsoft’s Azure endpoint.
Does anyone know if/when new models like this ( gpt-4-turbo-2024-04-09) get moved to Azure? I just got my account so am not sure how often new stuff ports over. Thanks!
Agree it would be useful to know.
In any case, I believe the following two pages are the key ones to follow for updates in this regard:
Perfect thank you! Yes looks like that What’s New page has model version updates from April 2024 so they keep it updated.
Yes, it would be incredibly beneficial to have access to the latest models. Additionally, there is no support for the retrieval tool on the Assistants API, which is surprising considering that OpenAI launched this tool several months ago… I am truly disappointed with the frequency of updates from Azure, almost to the point where I am considering not using their platform anymore.
It is tough to recommend, on one hand they give $125K in free credits to startups which is extremely generous… on the other hand, it’s difficult to set up (with confusing documentation, and weird unhelpful error messages around Quota).
I contacted their support too, and they seem to be at capacity in regions like USEAST which is a little mind-blowing considering their (Microsoft) involvement is to be the elastic compute resource… yet they have less capacity than the OpenAI API? Not sure I’d recommend but hopefully they start to improve things.
It seems Microsoft (and OpenAI) don’t really want to be in the business of selling shovels to miners, as the old adage goes that doesn’t consider the power of monopoly and regulatory capture. They want to be in the business of monopolizing the shovel market, acquiring and choking out those with shovel technology, making private shovel ownership difficult – and selling you a subscription to the gold in a bundle with other products.
