Claude AI doubles Claude Cowork usage limits for one month to support larger, more complex tasks

We doubled Claude Cowork usage limits for the next month. This applies to your 5-hr rate limits. If you’ve been saving up a big messy project, now’s the time.

Cowork is at its best on work that’s too big for a chat: research across dozens of accounts, recurring reports, triaging my inbox and drafting replies.

If you’ve been curious, this is a good month to find out what it can take off your plate. Can’t wait to see what you do.

In my experience, from Opus 4.6 onward, the bigger issue is not only usage limits. The model often answers too early and too confidently, so a lot of tokens get spent correcting, narrowing, and explaining what should have been handled with better uncertainty and scope control. Doubling limits helps, but reducing premature certainty would probably save more usage than simply adding more usage.Better self-limitation, clarification, and scoped answers matter too.

@claudeai Thank you. I just used Cowork to create this video:

@claudeai @hasanfr_0rg I always love using Claude Cowork, it’s very helpful.

@claudeai Your biggest users slice are the users in Claude Code, do you hate us? why we cannot have compensations like that for the weekend?

@claudeai best use case: - tree -L -2 -> infer relationships through filenames to cluster -propose cluster_ids - group files to cluster IDs or - do the packing_and_tagging for your project once - capture as skill - update the userFormFillRequest and execute many more times (readmes)

@claudeai I just cancelled my Max subscription. I wouldnt use it eve if it were for free.

It reached a point, where Opus 4.8 now is so dumb, its nothing more than a risk, to have this thing on your computer.

Also, the trust is gone too, who knows what kinda data Claude was sending home.

@Layton_Gott @IAM_KOPR8 @claudeai the funniest part of this is that not even claude agrees with you, you're just an asshole

@claudeai You’re going to have to do a lot more than that after the disaster that 4.7, 4.8 and everything completely being broken for a month has been. I’m literally using GROK to code because it’s more reliable than Claude now… 🤦🏻‍♀️

@MEM00063 @claudeai Same here, I always just say don’t share your thoughts or don’t be so verbose.

@claudeai Me for everything next month

@claudeai Cowo-wha?? never heard of it

@claudeai Literally will do anything but give your model the writing capabilities it had with 3.5/3.7

@claudeai what are you actually even supposed to do in cowork that you can't use http://claude.ai / claude code for

@IAM_KOPR8 @claudeai Yes but Claude Code can do that as well but better lol

@bcherny @VancityReynolds @TheRock @SidTheArgent @RyanGosling @AnaDeArmasHD 999 signal

@claudeai I’m confused, does this basically mean that the usage is only half the normal cost that it would be in cowork? Because otherwise, how do you double the usage for something specific like that when people are billed on total usage for the month across chat, coworker, and Codex?

@bcherny @VancityReynolds @TheRock @SidTheArgent @RyanGosling @AnaDeArmasHD and then he shows up

@bcherny Does this stack on the 50% limit increase already in place through July? I assume not