We’re having trouble connecting to the model provider. This might be temporary - please try again in a moment.
(Request ID: b0d18915-2c6a-40a5-a411-5de50e291cca)
1 Like
Yes, I can barely use any advanced model after the gemini downgrade event today.
I’m going to be stuck on 0.479 for life because I can’t afford to lose a day every two weeks when they ship. Edit & ask mode is fine for me if it works reliably.
I haven’t been able to use my company computer with cursor beyond 0.479 due to http2/proxy issues despite disabling http2 (which used to work).
Regarding the models being unavailable, the Gemini outage from around when you made your post caused some issues for us.
Almost out of nowhere, Gemini started erroring on the majority of requests. While we have systems in place for this, this then put a ton of demand on the other models, causing them to slow down and start to fail also.
We’ve learnt from this significantly, and a similar situation shouldn’t happen again.
Regarding @neuralengineer’s HTTP/2 issue, we should have a fix coming in the next week or two which should fix this for good. ZScalar (and similar corporate proxies) have some very weird behaviour, but we’re getting to the bottom of it currently and should have this substantially improved either before or in v0.51.