Anthropic’s response to Claude usage limit complaints only made users angrier

Anthropic apologized to frustrated Claude users today and explained why so many sessions were burning through limits faster than expected, but the response seems to have backfired.

Product lead Lydia Hallie posted the update on X, thanking people for their reports and saying the company had investigated. “We’re sorry this has been a bad experience,” she wrote. The thread racked up 1.3 million views in hours.

Hallie said the main reasons were tighter limits during peak hours and the higher cost of 1M-context sessions. The team fixed a few bugs, she added, but none were secretly over-charging users for tokens. They also rolled out efficiency improvements and added in-app pop-ups to stop big prompt cache misses.

Then came the tips that hit a nerve. For Pro plan users, Hallie recommended switching to Sonnet 4.6 instead of Opus 4.6 right at the start of a session because Opus burns through limits roughly twice as fast.

She suggested lowering the effort level or turning off extended thinking for lighter tasks, starting fresh sessions instead of resuming ones idle for about an hour, and capping the context window to save tokens. Anyone still seeing weird burn on small sessions should use the /feedback command, she said.

This update follows weeks of complaints we first reported on when Claude Max subscribers noticed usage limits draining with no clear reason. The team later tried to explain the situation, but subscribers weren’t buying it. Following that, a user alleged that two cache bugs sometimes made tokens cost up to 20 times more, which led Anthropic to investigate the claims. 

Still, a lot of subscribers aren’t satisfied with the investigation details. Many called the post tone-deaf and accused Anthropic of shifting the blame onto them.

Alex Volkov wrote that Anthropic was essentially telling people “you’re holding it wrong” by advising them to avoid Opus and 1M context on Pro plans.

Derya Unutmaz, MD, went further. He canceled his Max account and said it was the first time he had turned negative toward an AI company. “Such dishonesty!” he posted. “They claim to have the highest ethics & we need to trust them!?”

There are plently of other comments and reactions to the post, most of them being negative.

In case you missed it, this isn’t the only thing the company is struggling with. Just a few days ago, we reported on Anthropic’s Claude code source leaking, the follow-up DMCA takedowns on GitHub, and suspicions about a possible supply chain attack.

Anthropic says it’s still pushing more efficiency fixes and wants everyone on the latest version. Whether that will be enough after this week remains to be seen.

Featured image generated with AI

We stand out from the tech-media crowd because we break news stories; we mainly bring you stuff that you won’t find anywhere in the mainstream tech media. Our stories have been picked up by some of the world’s most popular websites and media outlets—more info is available here.