Early access — Free for Everyone
A neutral publishing layer for enterprise AI.

Your AI generated the report. Your company should own it.

Publish, govern and preserve AI-generated business knowledge independently from any AI vendor.

Works with Claude, ChatGPT, Gemini, Codex, and MCP-compatible agents.

From AI conversation to shared report.

The useful work already happens in your AI tool. publishwith.ai handles the step after: turning that output into a live link people can actually use.

1

Ask your AI

"Build the revenue dashboard and publish it for the exec team." No export, upload, deploy, or manual formatting step.

2

Get a live link

The report stays native: HTML, Markdown, charts, tables, and data-backed pages render in the browser instead of being flattened into a file.

3

Keep control

Update the same link, restrict who can open it, revoke public access, and keep a record of versions and views.

Built for people who turn AI output into business work

FP&A Analysts
Monthly revenue & forecast reports
RevOps Teams
Pipeline & churn dashboards
Consultants & PMO
Executive briefing packets
BI Teams
Data-backed operational reports
Team Leads
One latest version for stakeholders

The report is useful. The sharing is still broken.

AI can already produce the dashboard, briefing, or analysis. The mess starts when you need another person to open the result.

Slides and PDFs kill the output.

Charts stop being interactive, data goes stale, and the next update means rebuilding the package again.

Attachments create version chaos.

Stakeholders ask for the latest version, old files keep circulating, and nobody knows which copy is current.

Random links do not stay manageable.

Once a report leaves the chat, you still need updateability, revocation, access control, and a record of what was shared.

Use it where an AI report should become the source of truth.

"Analyze Q3 revenue and publish a board report. Exec team only."

Executive reporting

A board-ready link that stays current instead of a slide deck everyone forwards.

"Pull this week's numbers from our CRM and refresh the KPI dashboard every morning."

Live reports Team+

On Team and Enterprise, your AI can publish sandboxed Python that queries data and pulls approved external APIs, then re-renders on demand or on a schedule. The credential stays with your admin; the report just refreshes.

"Publish the Acme due-diligence report and send it to the Acme deal team."

Client & partner delivery

Keep a reusable group for each external client or partner, then direct every report you generate to them under access control, and revoke when the engagement ends.

Control is built into the link.

The first value is a report people can open and use. The second is that the same report can be restricted, revoked, versioned, and audited.

Corporate SSO

OIDC support for Azure AD, GitHub, Keycloak, and other identity providers.

Per-artifact ACLs

Owner, editor, and viewer roles for users and groups.

HTML sanitization

Uploaded HTML is sanitized and served with restrictive CSP headers.

Audit logging

Publish, view, access, and admin actions are logged with actor identity.

Guarded SQL

Agents can work with data while destructive operations stay classified and controlled.

Share links

Public read-only links are revocable, token based, rate limited, and tracked.

Encryption at rest

Every published file is encrypted with AES-256-GCM and a per-artifact key. Your content is protected even if storage is compromised.

GDPR-ready, hosted in Germany

All workspaces, databases, share links, and audit logs reside in the EU. No data leaves European infrastructure.

Audit evidence for AI governance

Exportable audit trail covering who published what, when, with which version, and who accessed it. The evidence GDPR and emerging AI Act oversight expect, ready to use inside your own framework.

For IT, this is governed publishing instead of another unmanaged link.

Business teams get a faster way to share AI output. IT gets a controlled distribution layer, not a new pile of shadow deploys.

1

Business users keep moving

They ask the AI for a report and receive a link. They do not need to learn hosting, deployment, or artifact management.

2

IT keeps visibility

Admins and owners can see what was published, what version is live, who has access, and which links can be revoked.

3

BI remains the system of record

Use publishwith.ai for AI-generated briefings, client packets, temporary dashboards, and narrative analysis around your existing tools.

Evaluating this for your organization? The governance page covers SSO, audit, access control, EU hosting, and security posture.

Governance for IT & security teams

What changes compared with the usual workarounds.

The difference is not another place to store files. It is a direct path from AI output to a live report link.

Outcome Email / Slack Shared drive Static hosting publishwith.ai
AI publishes directly - - - Yes
Report stays interactive - Partial Yes Yes
Same link can be updated - Manual Manual Yes
Revocable external sharing - Limited - Yes
Per-report access control - Limited - Yes
Data-backed dashboards - - - Yes
Version and access history - Limited - Yes
No rebuild into slides or PDFs - - - Yes
Early access — All features free until launch

Simple, transparent pricing

All features are unlocked during early access. Billing starts when we launch — early-access users get advance notice and a grace period.

Free

Personal projects and experimentation

€0/month
  • 1 user (single account)
  • 5 artifacts in your workspace
  • 5 MB total database (shared across artifacts)
  • 3 share links per user
  • Built-in auth & social login
  • HTML, Markdown, dashboards, and data-backed artifacts
  • MCP access
  • Default CDN allowlist: jsDelivr only
  • Live Python reports (sandboxed)
  • Full-text content search
  • Custom OIDC / SSO
  • Managed CDN allowlist
  • Account deleted after 30 days inactivity
Get started free

Enterprise

Enterprises with security and procurement needs

€199/month
  • Up to 200 users in your workspace
  • 500 artifacts in your workspace
  • 1 GB total database (shared across artifacts)
  • 200 share links per user
  • Multiple OIDC / OAuth2 providers
  • HTML, Markdown, dashboards, and data-backed artifacts
  • MCP access
  • Live Python reports — sandboxed, with governed API connections
  • Full-text search across all artifact content
  • Active as long as subscription is paid
  • Advanced CDN dependency policy
  • Security review, SLA, and priority support
  • Need more? Limits extendable on request
  • Custom subdomain (yourteam.publishwith.ai) Phase 2
Talk to us

On-premise & air-gapped

Self-hosted for enterprise

Run publishwith.ai entirely inside your own infrastructure. Full data sovereignty, your security perimeter, dedicated support and custom SLA. Tailored to your compliance and procurement needs.

Talk to our team

GDPR-ready. Hosted in Germany. All workspaces, databases, and audit logs reside in the EU.
Prices are indicative and may change before general availability. Early-access users will receive advance notice before billing begins.

Frequently asked questions

Is it free? What's the catch?
During early access, all features — including Team and Enterprise — are available to everyone for free. No credit card required. When we launch billing, all early-access users will receive advance notice and a grace period before any charges begin.
What do I actually publish?
Reports, dashboards, briefings, Markdown documents, data-backed pages, and small business artifacts generated by your AI. The result opens as a browser link, not as a flattened screenshot or attachment.
Do I need to know HTML or deploy anything?
No. You ask your AI to build and publish the report. publishwith.ai gives the AI a publishing target and gives you the resulting link, versions, access settings, and share controls.
Why not just export the report to PDF or slides?
PDFs and slides are useful for formal packages, but they flatten the AI output. Charts stop being interactive, data-backed pages stop updating, and every change creates another file. publishwith.ai keeps the report native and link-based.
Where is my data hosted?
All workspaces, databases, share links, and audit logs are hosted in Germany on EU infrastructure. The platform is built for GDPR compliance: data does not leave the EU, and you remain the controller of any personal data you process through your artifacts.
What is MCP and why does it matter?
MCP (Model Context Protocol) is how AI clients use external tools. For business users, the important part is simpler: Claude, ChatGPT, Codex, and compatible agents can publish the report for you instead of handing you files to copy, rebuild, or deploy.
Can I use my own identity provider (SSO)?
Yes. The Team plan includes one OIDC provider (e.g. Google Workspace, Okta, Azure AD, Keycloak). Enterprise supports multiple OIDC/OAuth2 providers. The Free tier uses built-in email/password auth plus social login (GitHub, Google).
Is each artifact’s database isolated from other artifacts?
Yes. Every data-backed artifact gets its own dedicated SQLite database file. There is no cross-artifact or cross-tenant data mixing. You can query it from Liquid templates, the JSON API endpoints, or browse it directly in the admin panel.
Can my AI build live reports that pull from our own data and APIs?
Yes, on the Team and Enterprise plans. Your AI can publish a live Python report: code that queries the artifact’s own database and pulls approved external APIs, then re-renders fresh output on demand or on a schedule. Because it runs code, it is off until a tenant admin opts in, and it runs in a hardened CPython-WASM sandbox — no raw network or filesystem, read-only database by default, and strict time, memory, and external-call limits. External APIs are reached only through named connections an admin defines: the host injects the credential behind an SSRF guard, so the script never sees the secret. The Free tier does not include Python reports.
Can I upload multiple files per artifact?
Yes. A single publish call can include as many files as needed — HTML, CSS, JS, images, fonts, and more. The full set of files becomes an immutable version. Asset size limits apply per plan: 50 MB total per version on Free and Team, 500 MB on Enterprise.
What happens to my artifacts when billing starts?
Nothing changes automatically. You will receive an email before billing begins. If you choose the Free tier, any artifacts beyond the 5-artifact limit will be archived (not deleted) and you can choose which ones to keep active.
What does “inactivity deletion” mean for Free accounts?
Free accounts with no login or artifact access for 30 days receive a warning email. If the account remains inactive for 3 more days, the account is deactivated and its workspace is taken offline. This only applies to the Free tier.
What happens to my Team or Enterprise account if I stop paying?
If a payment fails (expired card, declined charge), Stripe automatically retries for about 21 days. During this period you keep full access and receive email reminders. If the subscription is not recovered within that window, the workspace and all its data are permanently deleted. To preserve your artifacts, downgrade to Free before cancelling — up to 5 artifacts within Free limits remain active.
Why is jsDelivr the only external CDN allowed by default?
Published artifacts use a strict content security policy. By default, we allow one predictable CDN, cdn.jsdelivr.net, so AI-generated reports can use common libraries like Chart.js without opening every artifact to arbitrary third-party scripts. This keeps the security model simple to audit. Team workspaces can use an admin-managed CDN allowlist; Enterprise workspaces can define stricter CDN dependency policies through security review. You can also publish libraries as local artifact files.

Publish your next AI report as a live managed link.

Keep the output native, share one current version, and retain control when the report leaves the chat.

Publish your first report free