Changelog

The Quality build tier now runs on Claude Opus 5 by default - Anthropic's newest Opus, with noticeably sharper reasoning on multi-step build and edit work. New chats on Quality pick it up automatically; no setting to change.

Opus 5 thinks before it answers, so expect it to spend a little longer - and a few more credits - on hard tasks than the model it replaces. Claude Opus 4.8 is still one click away if you want the leaner option for everyday pages.

Switch any time in the Choose build quality dialog, or set your account-wide default in Settings -> AI defaults.

When you switch your project's shared link from "can edit" to view-only, everyone who already has the project open now loses edit access within a second - their editor switches to a read-only state with a clear notice. Turning editing back on unlocks connected visitors just as instantly.

Previously the toggle only applied to people who opened the link *after* the change: visitors who were already connected could keep typing until they reloaded the page, which made the setting look broken.

Opening a shared project link without signing in now gives you a way to browse the whole project, not just the files the owner happened to leave open.

The project menu has a new Files entry under View. It opens the file tree in a panel on the left - click any file to open it, exactly like the project's members do. Close the panel with the x in its corner and reopen it from the menu any time.

On read-only links the tree is view-only: you can open files, split the editor, and copy paths, but options that would change the project - creating, renaming, moving, or deleting files - are not shown. Links that allow editing keep the full set.

Kimi K3, Moonshot's flagship model, is now available in the Choose build quality dialog. It brings a 1M-token context window, always-on reasoning, and vision - so it can hold a very large project in context, think through multi-step changes, and read the screenshots or mockups you attach.

Pick it per chat like any other model, or set it as your account-wide default in Settings -> AI defaults. It is a strong fit for large builds where the whole codebase matters.

The Quality build tier now runs on Claude Opus 4.8 by default - our strongest all-round model for building and editing apps. New chats on Quality pick it up automatically; no setting to change.

Prefer a different model? Every option is still one click away in the Choose build quality dialog, and your account-wide default lives in Settings -> AI defaults.

The model you pick in the Choose build quality dialog now applies to that chat only. Open another chat, pick a different model, and run them side by side - one choice no longer follows you across every conversation.

Your account-wide defaults still live in Settings -> AI defaults, and the Set as my default row in the dialog now promotes your current models too, along with mode, effort, and context.

Published browser projects are now served from yourproject.playcode.site (previously on playcode.io). The new address shows up in the editor and dashboard, ready to copy and share.

If you shared an older playcode.io project link, visitors now see a short notice pointing them to the new .site address - reshare the new link wherever you posted the old one. Custom domains you've connected are unaffected and keep working as-is.

Converting a browser project to a Cloud full-stack app no longer fails when the project carries a large media library.

Projects with many hundreds of images could get stuck at the import step with an unhelpful error, and retrying or restarting the project computer never helped. The importer now transfers big asset sets reliably, and if anything does go wrong during a conversion, the error now names the actual cause instead of a generic failure - so the AI can fix it or route it to us with real detail.

Your signed-in home is rebuilt. Instead of a grid of files with a second, weaker prompt box, Home is now the place you build: a short greeting and the exact same composer you use inside the editor. From the very first screen you can attach files, use your voice, and pick your model and effort - no more stripped-down starter prompt.

• One composer everywhere. The box on Home and the box in the editor are the same tool. Whatever you can do in a chat, you can do the moment you arrive.
• Always know where your message goes. A small target chip on the composer shows whether your prompt will start a new project or continue an existing one. If you have a single project, it points there by default; one tap switches it - so a quick change request never spawns an empty project by accident.
• A project sidebar on every screen. Your projects, their chats, and folders now live in one collapsible sidebar. Open any project or jump straight to a past chat in a single click, drag a project into a folder, and reorder folders to taste. It stays put as you move between Home and the editor, so you never lose your place.
• A template gallery when you want it. "Or start from a template" opens a browsable gallery of ready-made site templates, with coding templates - React, Vue, plain JavaScript, and more - one click away. It is opt-in and never gets in the way of just describing what you want.
• A dedicated Archive. Archived projects have their own page with bulk restore, keeping your active project list clean.
• An editor that earns its chrome. A brand-new project opens clean and quiet. Share, Publish, and other controls appear as your project becomes real - once you have actually built something worth sharing - so the toolbar teaches you the workflow instead of overwhelming you on day one.

Old links to your dashboard bring you straight to the new Home, so nothing you had bookmarked breaks.

Cloud publishing got a stronger guarantee: Publish deploys the exact code that is in your workspace at the moment you press the button - every time, no matter what the AI has been doing with Git behind the scenes.

Previously, if the AI organized its work into a Git branch, a publish could quietly ship an older version of your app, or fail repeatedly with a cryptic Git error that retrying never fixed. That whole class of problems is gone:

• What you see is what ships. Publish captures your current workspace state as a single immutable version - unsaved edits included - and deploys exactly that. Branches, background commits, and Git bookkeeping can no longer change what goes live.
• Every publish shows its source. The publish dialog and history now identify each attempt by its branch and version - so "what exactly is live right now?" always has a clear answer, for you and for the AI.
• Honest errors instead of retry loops. In the rare case your workspace is genuinely out of sync with the published history, Publish stops before touching production and explains what happened in plain words - your live site stays untouched, and the AI can see the same explanation and fix the mismatch for you.