What exactly did Anthropic make faster?

Anthropic announced it sped up the user experience of the web service claude.ai and its desktop app, by about 3 times in 2 weeks.

Who did the work of speeding it up?

Not only human developers. Anthropic says Claude itself found the points that needed improving and repeatedly corrected them.

It is a development support tool that handles not just writing code but also investigating and fixing existing code and running tests. In some reported cases it pursues a given goal on its own without a human directing each small task.

Large Language ModelsAI Business & IndustryITmedia AI＋3m agoOct 2, 2026, 10:01 JST

Anthropic: claude.ai 3x faster in 2 weeks, with Claude's own fixes

Send on LINE

3 Key Points

What happened
Anthropic said it made the web service claude.ai and its desktop app about 3 times faster in 2 weeks, and that Claude itself found and fixed the points to improve.
Why it matters
The speedup came not only from human developers but from Claude reviewing and correcting its own work, pointing to a development approach that may need less step-by-step human instruction.
What to watch
The body gives no benchmark, no baseline, and no date for the claim, so the result hinges on how Anthropic measured it. Watch whether the same approach spreads to other products.

WHO IT HITSThis lands on product and engineering teams that ship AI chat apps and desktop clients, since the reported gain came in part from having the AI find and fix its own weak points rather than from developer effort alone.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, in your inbox every morning.

Get news like this every morning, free

Context & Analysis

The announcement sits alongside a broader shift described in the article: Anthropic's development support tool, Claude Code, now handles not only writing code but also investigating and fixing existing code and running tests, and there are reports of it pursuing a given goal on its own instead of waiting for humans to direct every small step. The speedup of claude.ai and the desktop app is presented as another instance of that direction, with Claude taking part in locating and correcting its own weak points.

The article frames this as a slightly unexpected development method — not simply having AI write code, but having the product contribute to its own improvement. Beyond the headline figure, the body gives no baseline, benchmark, or measurement method, so how large the gain really is remains something Anthropic has yet to spell out. Whether this self-correcting approach becomes routine for other products is the open question.

FAQ

What exactly did Anthropic make faster?: Anthropic announced it sped up the user experience of the web service claude.ai and its desktop app, by about 3 times in 2 weeks.
Who did the work of speeding it up?: Not only human developers. Anthropic says Claude itself found the points that needed improving and repeatedly corrected them.
What is Claude Code?: It is a development support tool that handles not just writing code but also investigating and fixing existing code and running tests. In some reported cases it pursues a given goal on its own without a human directing each small task.

ITmedia AI＋Read Original Article

Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Get it free

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.