mdedit.ai
mdedit.ai is a Markdown workspace for technical content, pairing a long-document editor with reviewable AI edits, live collaboration and direct publishing. It runs in the browser or fully offline on desktop, with optional local models.
What is mdedit.ai?
mdedit.ai is a Markdown workspace built for people whose job is technical content: documentation, READMEs, long-form guides and internal notes. Its premise is stated plainly on the homepage as AI drafts, you approve — the assistant proposes, every change appears as a visible diff, and ordinary human comments keep the final call where it belongs.
The product is organised around five pillars. The editor keeps source and rendered output side by side, lints and autofixes formatting drift, generates a table of contents for long documents and preserves frontmatter. The context-aware AI works with the open document in view through Ask, Edit, Create and advanced Agent plans, and feeds an AI Review into the same comment threads a colleague would use; models come from OpenAI, Anthropic, Grok or Gemini, via included credits, your own key, or built-in access on the top tier. Reliability and sync add automatic version history with compare and restore, offline-first editing that survives a network drop, and local Markdown workspaces on desktop that stay aligned with files changed outside the app. Visual content covers generated photos, diagrams, charts and animations alongside hand-authored Mermaid, Graphviz and Markmap, a visual table editor with CSV import, and presentation-ready code images. Review and publish brings live co-editing with presence, comments, suggestions, assignments and resolution, granular permissions, hosted links, and one-step publishing to GitHub, WordPress, Ghost, DEV, Hashnode, X Articles, Gists, S3 or any webhook.
It is deliberately multi-surface. There is a web app, a desktop build for macOS, Windows and Linux — version 0.7.0, published in June 2026 — that can run fully local models through Ollama or any OpenAI-compatible server, native Android and iOS companions, browser and VS Code/Cursor extensions, an npm CLI and a hosted Model Context Protocol server for agents. A scratchpad opens with no sign-up at all. The site claims more than 8,800 users and 5,000 documents edited, and credits a single creator, Vivek Maskara, with building the focused workspace he wanted after years of fragmented publishing workflows.
What it does
- Write and structure long Markdown documents with live preview, linting and a generated table of contents
- Ask AI to draft or rewrite with the document in context, then accept or reject each change as a visible diff
- Co-edit in real time and run review through comments, suggestions, assignments and resolutions
- Create technical visuals: Mermaid, Graphviz and Markmap diagrams, tables, code images and generated charts
- Publish the same source to GitHub, WordPress, Ghost, DEV, Hashnode, X Articles, Gists, S3 or a webhook
- Export to Markdown, HTML, DocX, ODT, EPUB or PDF, and import Markdown, HTML or Jupyter notebooks
- Drive articles from the terminal with the CLI, or from an agent through the hosted MCP server
When to use mdedit.ai / When not to
A quick filter to help you decide if mdedit.ai is the right fit.
When to use mdedit.ai
- Technical writers who own long documents from first draft through subject-matter review to publication
- Software developers writing READMEs, architecture guides and developer education in Markdown
- Technical content teams coordinating writers, SMEs and publishers around one live Markdown source
- Students taking notes, reviewing group work and exporting papers, reports or a Markdown resume
- Privacy-sensitive writers who need offline drafting and fully local AI models on the desktop app
When not to use mdedit.ai
- Writers who want a WYSIWYG word processor rather than a Markdown-first editor
- Teams building a linked knowledge graph, a contrast the site itself draws with Obsidian
- Mobile-first users: the mobile apps still lack publishing, repository and advanced editor surfaces
- Free-tier users needing more than three cloud-synced articles, or any collaboration feature
- Enterprise buyers who require a Data Processing Agreement, security certifications or a published legal entity
How to use mdedit.ai
A typical end-to-end flow, from setup to results.
- Open the scratchpad and edit a real document without creating an account
- Create a free account when you want your work synced; no payment card is required
- Organise the work into workspaces, folders and subfolders, moving articles in bulk if needed
- Draft in Markdown with live preview, linting and the generated table of contents beside you
- Choose your AI route in Settings, AI: included credits, your own provider key, built-in AI, or a local Ollama server on desktop
- Call the assistant in Ask, Edit, Create or Agent mode and review each proposal as a diff before accepting it
- Invite collaborators with read or write access, then work through comments, suggestions, assignments and resolutions
- Add visuals: Mermaid, Graphviz or Markmap diagrams, a CSV-imported table, or a generated illustration
- Publish to a connected destination, or export the file in whichever portable format you need
- For automation, install the CLI with npm, create a scoped API key in Settings, Integrations, and register the MCP server for agent workflows
Pros & Cons
Pros
- Markdown stays the source of truth, so documents remain portable and easy to take elsewhere
- AI proposes rather than imposes: every change is a diff you accept or reject, never a silent rewrite
- Genuine offline work, with local Markdown workspaces and local models that keep documents on your machine
- One place for drafting, review, visuals and publishing instead of four disconnected tools
- Low entry cost: a permanent free plan, then USD 6.99 per month, and up to 52% off with annual billing
- Unusually broad surface coverage: web, desktop, mobile, browser and VS Code extensions, CLI and MCP server
- A detailed privacy policy that spells out what is deliberately not tracked, including document content
Cons
- No postal address, no incorporated company and no declared country of establishment anywhere on the site
- No Data Processing Agreement, no security certification and no EU Article 27 representative
- Nothing is stated about whether customer content may be used to train models, and no opt-out is documented
- The free plan is tightly capped at three cloud-synced articles for the life of the account, with no collaboration
- Liability is capped at the greater of twelve months of fees or one hundred Canadian dollars
- Mobile apps are documented but no App Store or Google Play listing is published, and they lack publishing features
- No enterprise tier, no documented single sign-on, and the interface appears to be English-only
Pricing & Plans
A permanent free plan is available at no cost. The cheapest paid tier is Standard at USD 6.99 per month, or USD 39.99 per year; Premium is USD 17.99 per month, or USD 129.99 per year. Both paid plans include a seven-day free trial, which requires a payment card, whereas a free account does not. Payments are processed through Stripe, and fees are stated as non-refundable except where the law requires otherwise.
- core Markdown editor
- 20 trial AI credits
- 3 cloud-synced articles for the lifetime of the account
- 3 version snapshots per article
- export to Markdown
- HTML
- Word and PDF
- unlimited cloud-synced articles
- AI with your own key
- real-time collaboration and shared editing
- inline comments
- suggestions and highlights
- 50 version snapshots per article
- 3 team workspaces
- standard support
- everything in Standard plus built-in AI with no API key
- advanced AI models
- unlimited version history
- advanced collaboration and priority support
Data, GDPR & hosting
A consolidated view of how mdedit.ai handles your data.
GDPR overview
Implementation is concrete but partial. The privacy policy carries a dedicated Legal Basis for Processing section naming four grounds with their articles: contract performance 6(1)(b), legitimate interests 6(1)(f), consent 6(1)(a) and legal obligation 6(1)(c). A section for EU and EEA residents lists seven rights: access, rectification, erasure, restriction, portability, objection and withdrawal of consent. The controller is named, a Data Protection Officer is designated, and both are reachable at contact@mdedit.ai. Deletion is promised within thirty days. The gaps are just as explicit: no EU representative is designated under Article 27, no Data Processing Agreement is published or offered on request, no security certification is claimed, and analytics and advertising cookies rely on implied consent from continued browsing. International transfers, including to the United States, are announced with unspecified safeguards.
Who owns the data?
The Terms of Service state plainly that you keep ownership of the Markdown documents, files and comments you submit. Mango App Studio receives only the limited rights it needs to host, process, back up, transmit and display that content, and solely to operate, secure and improve the service or carry out the actions you request. Nothing grants a broader licence. Documents stay private until you or an authorised collaborator changes their visibility. You remain responsible for holding the rights to whatever you upload, edit, share or publish. The privacy policy adds that personal information is never sold to third parties.
Reuse rights
Because ownership stays with you, your documents can be reused, relicensed, republished or archived freely, with no permission to ask. Export is treated as a right rather than a favour: Markdown, HTML, DocX, ODT, EPUB and PDF are all available, and the CLI can pull raw Markdown or a ZIP archive straight into a pipeline. Publishing pushes the same source to GitHub, WordPress, Ghost, DEV, Hashnode, X Articles, Gists, S3 or any webhook you configure. One limit deserves attention: unpublishing disables the mdedit.ai public link but cannot recall copies other people have already saved or shared. Content you send to an AI provider or to a connected agent host is then governed by that third party's own terms and retention rules.
Data retention & training
Hosting summary
The privacy policy names AWS and Vercel as hosting providers but never states a hosting country or region. The only geographic commitment is a transfer clause: information may be transferred to and processed in countries outside your country of residence, including the United States, with appropriate safeguards that are not described. The domain resolves to an anycast address on Amazon infrastructure, which locates a network node rather than a datastore. Stated security measures are encryption in transit over HTTPS/TLS, encryption at rest for sensitive data, hashed passwords, regular audits, access controls and monitoring. Payments run through Stripe, analytics through PostHog with Google Analytics 4 as a secondary tool. Two things soften the picture for privacy-sensitive users: browser drafts are saved locally first and sync afterwards, and the desktop app can work directly on local Markdown files with a local model, in which case documents and prompts never leave the machine. The hosted MCP service validates tokens per request and stores neither access nor refresh tokens. No jurisdiction choice, data-residency option or Data Processing Agreement is offered.
Things to keep in mind
Risks and trade-offs to weigh before adopting mdedit.ai.
- Approving AI suggestions becomes a reflex: the diff view only protects you if you genuinely read it, and attention erodes over a long review session
- Leaning on AI drafting for documentation can quietly dull the harder skill of explaining a system in your own words
- Publishing is one click and effectively irreversible: unpublishing kills the link but not the copies others have already saved
- The publisher is a single individual trading under a business name, with no published address, no legal entity and liability capped at one hundred Canadian dollars
- Nothing on the site says whether your document content may be used to train models, and no opt-out is documented
- On paid accounts your email address is attached to analytics as a person-level property, and tracking spans both the marketing site and the app
- The free plan's three-article lifetime cap can strand work you assumed was safely synced to the cloud
Setup & Integrations
Technical difficulty
Very low for the web app: the scratchpad opens without an account, and a free account needs no payment card. The desktop app is a standard download for macOS, Windows or Linux with an install guide. Difficulty rises only for optional surfaces, such as pasting a provider API key, installing Ollama or LM Studio for local models, running npm install -g mdedit-cli and authenticating with a scoped API key, or registering the MCP server with a single npx command. Nothing requires servers, infrastructure or configuration files.
Deployment
Integrations
Behind mdedit.ai
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement mdedit.ai.
Frequently asked questions
Is mdedit.ai free to use?
Which AI models can I use?
Do I need a credit card to try the paid plans?
Can several people edit and review the same document?
Does mdedit.ai work offline?
Where can I publish, and what can I export?
Is there an API or an agent integration?
How do I delete my account and my data?
Should you pick mdedit.ai?
mdedit.ai is unusually complete for a product built by one person. Long-document editing, reviewable AI, real-time review, diagrams, version history, local files, a CLI and an MCP server together cover the whole path from first draft to published guide, and they do so without ever taking Markdown away from you. Every product surface we probed responded, the documentation is dense and current, and the free scratchpad lets anyone judge the editor on a real document before creating an account.
The pricing is genuinely low. A permanent free plan exists, the first paid tier is USD 6.99 per month, and annual billing cuts that by 52%. Bring-your-own-key AI from Standard onwards, and fully local models through Ollama on desktop, make this one of the few writing tools where confidential drafts can stay on your machine at no per-token cost.
The reservations concern the publisher rather than the product. mdedit.ai is operated by an individual, Vivek Kumar Maskara, trading under the Mango App Studio name. No postal address, no incorporated company and no country of establishment are published, and the terms cap liability at the greater of twelve months of fees or one hundred Canadian dollars. There is no Data Processing Agreement, no security certification, no EU Article 27 representative, and no statement at all about whether customer content may be used to train models. The free plan's three-article lifetime cap is also tighter than it first appears.
For an individual writer, a developer or a small technical-content team this is an easy tool to recommend, and an easy one to leave, since the files remain yours. For a regulated or enterprise purchase, the missing paperwork should be settled before signature.
- Choosing a selection results in a full page refresh.
- Opens in a new window.