Charming
Charming is a collaborative cloud for apps built with Claude, ChatGPT, and other AI assistants. It hosts the apps, persists their data, and lets users share them with controlled permissions so the apps can be used in chat, in a browser, or on a phone.
At a Glance
- Individual consumers building personal apps
- Primary-school educators and classrooms
- Consultants and freelancers
- Small companies and collaborative teams
- +1 more
AI Tools by Charming
(1)
Charming
Hosting for AI Built Apps
Discussions
No discussions yet
Be the first to start a discussion about Charming
Latest News
Charming MCP repository synchronized its tool list from the live MCP contract.
Automated MCP facts/tool-list synchronization was merged into the Charming repository.
Charming completed its official MCP Registry publication work; the repository notes the server was already registered under the tambo-ai namespace.
The repository added an Agent Plugins v1.0.0 package target and replaced its schema validator with Ajv.
Products & Services
A hosted runtime for apps generated by AI assistants. Charming publishes each app at a live URL, provides per-app storage and persistent state, and supports private sharing with viewer, end-user, collaborator, and related permissions.
A remote MCP server at https://charm.ing/mcp (and a ChatGPT-specific endpoint) that lets compatible AI clients create, render, update, query, and operate hosted interactive web apps.
An HTTP integration path for agents or clients that cannot use MCP, allowing them to create and operate Charming apps and access the same persistent data.
Market Position
Charming positions itself between chat-native AI artifacts and conventional developer hosting. It differentiates from Claude Artifacts and ChatGPT Canvas with a neutral live URL, persistent database-backed state, cross-assistant access, and granular sharing; versus Lovable, Replit, Bolt, and v0 it emphasizes building through the user's existing AI assistant; versus Vercel and Netlify it provides managed storage, authentication/sharing, permissions, and deployment without technical setup.
Leadership
Founders
Michael Magan
Founder of Charming; spent 10 years building products at Indeed, Convoy, and Taxbit. He previously built an open-source generative-UI project that the company says has more than 10,000 GitHub stars.
Executive Team
Michael Magan
Founder
Product builder with prior product experience at Indeed, Convoy, and Taxbit; also built an open-source generative-UI project.
Lachlan Heywood
Head of Engineering
Has worked in open-source cloud and distributed infrastructure; most recently was UI lead at Cosmonic.
Founding Story
Charming began at the start of 2026 after one of the engineers returned from family leave with an app for tracking a baby's feedings. The team kept seeing non-technical people build small useful apps, but found existing hosting platforms were not a good fit; Charming was started to give those apps a simple home with persistent data and sharing.
Business Model
Revenue Model
Freemium subscription SaaS: free personal apps, paid monthly Pro and Team subscriptions, and custom-priced Enterprise contracts. Paid tiers monetize increased storage, runs, version/history capacity, custom domains, collaboration, governance, and enterprise identity features.
Pricing Tiers
Unlimited apps, 10 saved versions per app, 1,000,000 runs per month, and 1 GB storage.
Custom domains, removes the Charming mark, 50 saved versions per app, 5,000,000 runs per month, and 25 GB storage.
Everything in Pro plus shared apps, admin controls for shares/APIs, audit trail, 5,000,000 runs per seat per month, and 25 GB storage per seat.
Team features plus single sign-on, roles and permissions, an admin view of every app, and contract-specific limits and terms.
Target Markets
- Individual consumers building personal apps
- Primary-school educators and classrooms
- Consultants and freelancers
- Small companies and collaborative teams
- Fortune 100/company employees and enterprise IT teams
- Personal software and small utilities built by non-technical people
- Teachers and primary-school classrooms
- Consultants and freelancers building tools for clients
- Employees at large companies building internal tools with AI
- Small teams sharing and operating apps across multiple AI assistants
- Turning Claude Artifacts or ChatGPT-built prototypes into persistent standalone apps
- Primary-school teachers
- Consultants and freelancers
- People at Fortune 100 companies