thanks for the feedback on the Claude Code Desktop launch this week. we'll continue to iterate based on feedback (yes, group by folder is coming back). our small team is grateful to be building for so many of you!

how much does it cost @elonmusk?

it all begins with a flower

debating following @theo's lead. should we switch to group by status by default?

🌟 Claude Code on the desktop got a serious glow up. It's honestly way better. Try these 7 new features: 1. Saved environments: save a few folders, an SSH host, or cloud as one setup and start sessions there 2. A worktree per folder, so parallel sessions don't step on each other 3. Nested sessions: ask for parallel tasks and each one reports back to the main thread 4. /btw (Cmd+;): side questions open in their own tab while the main session keeps working 5. One-click follow-ups: Claude suggests the next task, one click runs it as its own session 6. Sidebar status: see which sessions are working and which need you 7. Recap cards + a live to-do list, so you can catch up on a session in seconds The new /btw is probably my favorite. Hard to say, these all sound great. How about you?

we've been working on simplifying the ui and making it easier to manage multiple sessions. there are a lot of things in this update, hope you enjoy it!

wish i worked on this tbh. it's cool how fast our tech is changing.

I don’t want to manage a agents with different personas. If intelligence is truly super there should be no need for roles. Smart generalists rule the world. If you’re making agents with personas, are you actually finding value from it?

building is for everyone

Humanity’s reach expands as we innovate on the tools we use. Tools like Claude Code enable people to do things that once seemed impossible. 🤯

ICYMI we've cut the price of Sonnet 5.5 cache reads in half. In the Claude Platform, it's now $0.10 per million tokens (input is $2, output is $10). This means Sonnet 5.5 now runs ~20% cheaper on most agentic work. (API only, no change to Claude Code usage limits.)

"We may hope that machines will eventually compete with men in all purely intellectual fields." - Alan Turing (1950)

really excited to see how folks use Haiku to bring more delight to their products!

claude where you work

Organize your tools your way in CC desktop!

tag me on all your desktop mods 👀

Would love to learn if you're not using low as the default coordinator effort level in Projects and why!

ok, just saying it outright: do not use Sonnet with max effort! at that point, you should probably be using Opus Sonnet is the model that balances quality/speed/cost so when you push it to max, it thinks a lot more... spends more time... and costs more. you start to lose what makes Sonnet useful! this is why we have default effort levels, and in Claude Code the default is medium. all the effort dials are still there for anyone that needs them for an edge case. but our recommendation for most is: don't touch that dial!

everyone was really enjoying opus 5.5 last week so we thought we'd bring you more fun this week.

Claude Opus 5.5 (High) debuts at #1 in Text Arena with 1509 pts! That’s an 18-pt improvement over Opus 5 (High), now at #11. Opus 4.6 (High) remains in the #2 spot, just 4 pts off the lead. This release gives @AnthropicAI all six top spots in Text Arena! Opus 5.5 (High) also landed on the Text Arena Pareto frontier at a blended $16/MToken, based on input/output pricing per million tokens! See its placement in the post below. Congrats to the @AnthropicAI team on the release!