A URL for every artifact Claude makes
Where your teammates
and Claude work on artifacts.
One command from Claude Code, Claude Desktop or the web app. Your team opens it with the Google or Microsoft account they already use – no Claude seat, no download, no expiring link.
The demo shows an agent publishing a doc to display.dev – from a chat or a terminal – then the team opening its URL, commenting, the agent publishing v2 at the same URL, and the doc being shared with someone outside the company.
Cursor, Codex, Claude Code – any agent that writes a file. Switch agents anytime – your URLs, versions and comments stay put.
Indigo Engineering
Anthropic's share link isn't a URL for your team
Claude's built-in publish gives you a claude.ai/public/artifacts/… URL. Fine for a public demo. Claude Code Artifacts add a private page on claude.ai, but on Pro and Max it stays private to you alone – sharing is Team and Enterprise only, it reaches authenticated seats inside your Claude org and nobody else, the source has to be Claude, and there are no comments.
display.dev opens viewing to anyone with their company sign-in and no Claude seat, adds comments and updates from any agent, and outlives the chat window.
- ✗Company data on the open webA public URL is the only way to show an outsider the page.
- ✗Feedback goes to SlackNo inline comments, so notes land somewhere the agent can't read.
- ✗The chat expires, the link should notSix months later the context is gone and so is the artifact.
- ✗Fifty artifacts, no addressesNothing to organize, nothing on your own domain.
- ✗Another consumer share surfaceOne more thing for your security team to say no to.
Three ways to publish from Claude
Publish it with one command
dsp publish ./artifact.html
# → https://acme.dsp.so/0kzNYG7O-revenue-dashboardThe CLI uploads the file, returns the URL and copies it to your clipboard. What opens is the artifact itself – JavaScript, charts, filters, hover states – not a screenshot of it.
To change it later, republish with the artifact's short_id and current base_version. Same URL, new version, comments and history intact.
Who can open it
The same shape as Anthropic's publish, on your own domain, with comments and version history added.
Named email addresses only. Useful for a client or a vendor you don't want to hand company sign-in to.
Anyone signing in with a verified company domain gets in – Google Workspace, Microsoft Entra ID, or a one-time password by email. Viewers are unlimited and free.
You change the level any time. Move an artifact from company-gated to public and back. The URL stays the same.
What Claude builds, your team opens
Anything Claude can write, your team can open in a browser. Six artifacts published from real agent runs – every one of them live right now.
Compared to the alternatives
| Method | Its own URL | Company auth | Invited people | Inline comments | Custom domain |
|---|---|---|---|---|---|
| Anthropic publish | ✅ (on claude.ai) | ❌ | ❌ | ❌ | ❌ |
| ShareMyArtifact | ✅ | ❌ | ❌ | ❌ | ❌ |
| ShareDuo | ✅ | ❌ | ❌ | ❌ | ❌ |
| OneClickLive | ✅ | ❌ | ❌ | ❌ | ❌ |
| Tiiny Host | ✅ | ❌ | ❌ | ❌ | Pro |
| display.dev | ✅ | ✅ | ✅ | ✅ (read via MCP) | Enterprise |
The other tools stop at "the link works." display.dev is built for what happens after that: humans leave feedback, the agent reads it back and republishes.
Publish, gate, comment, iterate – at one URL
Claude publishes, your team comments, then it republishes with the artifact ID and base version.
Pricing
Try it. No credit card.
- Included: 10 gated artifacts
- Included: Unlimited public artifacts*
- Included: 50MB storage
- Not included: Version history
- Not included: Google + Microsoft SSO
- Not included: Audit logs
display.dev branded
Individuals. Unlimited gated sharing.
- Included: Unlimited gated artifacts
- Included: Unlimited public artifacts
- Included: 1GB storage
- Included: 10 versions
- Not included: Google + Microsoft SSO
- Not included: Audit logs
For teams that need company auth.
- Included: Unlimited gated artifacts
- Included: Unlimited public artifacts
- Included: 25GB storage
- Included: 50 versions
- Included: Google + Microsoft SSO
- Included: 90 days audit logs
For compliance-driven organizations.
From €499/mo
- Included: Unlimited gated artifacts
- Included: Unlimited public artifacts
- Included: Custom storage
- Included: Unlimited versions
- Included: Google + Microsoft SSO
- Included: 365 days audit logs
Free covers 10 gated artifacts, unlimited public links and unlimited viewers – enough to run the loop with your team before you pay anything.
Features
A PM highlights a paragraph and drops a note. Your agent reads the thread over MCP, fetches the current version, republishes with short_id and base_version, then resolves the thread.
Google Workspace and Microsoft Entra ID, or a one-time password by email. No new accounts for viewers, no per-viewer license.
Every republish adds a version to the same link. Viewers always get the latest; earlier versions stay one click away on the version pill.
Claude Code today, Codex or Cursor tomorrow. The URLs, versions and comments stay where they are.
Interactive charts, custom CSS, embedded data. If Claude can write it, display.dev serves it.
Who opened the artifact, when, and which sign-in they used – on Pro and above.
What teams say
It's helped us delay a switch to Anthropic enterprise plans, which we probably would've been forced into so we could collaborate on Claude Design output.
Try it free
- Publish a Claude artifact in under a minute, no card
- Keep the URL when you upgrade
- Unlimited viewers on every plan
Common questions.
Does this work with Claude Code, Claude Desktop and the web app?
All three. Claude Code runs dsp publish in the terminal. Claude Desktop publishes through the MCP server. For an artifact or Claude Design prototype built in the claude.ai web app, export its standalone HTML file and publish it in app.display.dev. In three exports we verified, the interactions we exercised still worked after publishing: a screen change, FAQ expansion and search filtering.
Do viewers need a Claude seat or a display.dev account?
Neither. They open the link and sign in with the Google or Microsoft account they already use, or with a one-time password by email. Viewers are unlimited on every plan, including Free – you pay for publishing, not for the people reading.
Can Claude read the comments back and update the artifact?
Yes, and it's the reason the page exists. The MCP server exposes the comment threads alongside get, publish and resolve. Claude reads a thread, fetches the current version, edits, republishes with short_id and base_version, and resolves the thread – no human in the middle for the maintenance pass.
Does the URL stay the same when Claude republishes?
Yes. Each republish adds a version to the same link, so anything you pasted into Slack, Notion or a calendar invite keeps working. Viewers get the latest; earlier versions stay one click away on the version pill.
Does the artifact still work, or does it turn into a screenshot?
It works. What you publish is the file itself, so JavaScript, CSS, charts, filters and hover states all behave exactly as they did in Claude. HTML and Markdown both render.
How big can an artifact be?
Up to 10 MB per file on Free and 50 MB on paid plans – far above anything Claude writes as a single page, including artifacts with data embedded in them.
I have HTML that Claude didn't write. Does that work?
Yes. display.dev never inspects what produced the file. Same flow for Cursor, Codex and v0, and sharing a plain HTML file with your team works the same way.
Can I use a custom domain instead of `*.dsp.so`?
Yes, on Enterprise. Bring your own domain (e.g. share.acme.com), point a CNAME at us, done.
Publish your first artifact in 15 seconds.
Free tier. No credit card. One-time password auth on free, Google + Microsoft SSO on Pro (€49/month flat).