Over the last week or so, I frequently run into Claude not responding with the message below. I have a paid account, and I have not been overusing it. This seems to be systemic. As a result, I am using ChatGPT and Gemini more often despite my qualms about Open AI my preference for Claude when it comes to editing writing.
Has anyone else been experiencing similar issues with Claude?
They all go through their phases. May also depend a bit on where you are serviced from, service interruptions can be restricted to a locality sometimes. I am on the max/200 plan i.e. I use claude many hours a day and have not experienced an elevated lack of service level. Guesstimating I would say I have had a maybe 1 hour collective outage for the last month.
Reliability hasn’t been as good as usual recently. Claude was unusable for me for over an hour this morning, but it seems to be working again for now.
I’ve been keeping an eye on Claude’s status and have subscribed to be notified about issues affecting Cowork.
you are going to be inundated with ‘ping’ :-D. There are usually many little ‘issues’ that normally you’d not notice were it not for that notification. Drove me nuts so its off for me.
I have had some instances. Usually, I drop down to a lower model and it connects.
Thanks for the heads up, @Joost. I purposely chose to be notified only about Cowork issues, but I will unsubscribe if there are too many notifications and only resubscribe if there’s another outage.
I’ve had similar availability issues with Claude. Worse, Claude’s answers seem to be more flimsy than usual using Opus 5, even on the “Max” thinking mode.
I’m considering a (temporary) switch to OpenAI’s ChatGPT as it’s been about a year since I’ve tried any of the other frontier model providers.
I have had the same experience with Claude’s responses of late. Consequently, I have been using ChatGPT more. At least for now, I find I prefer ChatGPT’s responses to Claude’s. I find Gemini to be the worst of the three.
I’ve found that I often get better results if I dial Opus 5’s effort levels down rather than up. It’s almost as if it ties itself in knots in overthinking when the effort level is higher than it needs to be. Sometimes it over-explains itself at higher effort levels, too, and wanders into thickets of throat-clearing and table-setting.
Increasingly lower quality for me. It’s like every time they come up with a new model the previous one gets dumber. Combined with the fact that the new one (fable) eats tokens faster and they are removing the “free” 50% they have been “giving” us in a few days I’ll be moving to other providers as well.
