
Miro
Miro is an AI innovation workspace: an infinite visual canvas where product, design, engineering and marketing teams run workshops, diagrams and roadmaps side by side with AI agents, AI workflows and their own connected AI tools.
What is Miro?
Miro presents itself as an Innovation Workspace: an infinite, multiplayer visual canvas where teams and AI agents work on the same surface. It began in 2011, when Andrey Khusid needed a way to show ideas to remote clients of his design agency; the first version, built with co-founder Oleg Shardin, was called RealtimeBoard. The company now claims more than 100 million users and 250,000 organisations, and names Salesforce, Asos, Deloitte, WPP and Cisco among its customers. Its 2022 Series C release described it as co-headquartered in San Francisco and Amsterdam, and it lists thirteen offices, from Berlin and Paris to Singapore, Tokyo and Yerevan.
The product is organised around five blocks: the Intelligent Canvas; native formats covering Docs, Tables, Slides, Diagrams, Kanban and Timelines; Blueprints for repeatable processes; enterprise security and scale; and integrations. On top sits the AI layer. Sidekicks are contextual AI agents, taken from prebuilt libraries or customised, that build on the canvas. Flows are multi-step visual AI workflows that can be turned into templates and shared across an organisation. Connectors bring live data from third-party tools onto the board, and an MCP server links Miro to external AI tools in both directions: generating code from a board's content, or creating diagrams and documents from inside the AI tool. The idea the company pushes hardest is to use your entire canvas as a prompt, so that everything a team has built becomes context. Nine AI Playbooks package that into starting points for planning and strategy, product development, rapid prototyping, competitor analysis, process transformation, research synthesis, technical architecture, brand campaigns and service blueprints.
AI usage is metered in monthly credits, pooled at organisation level, consumed per AI action and never carried over. Facilitation features complete the set, with TalkTrack for asynchronous narration and Engage for polls, 2x2 matrices and scales. Miro displays ISO 42001, the AI management standard, alongside ISO 27001, SOC 2, TISAX, NIST and GDPR compliance. Recent moves back the repositioning: the MCP server launched in February 2026 with Anthropic, AWS, GitHub, Google and Windsurf, Reforge was acquired in March 2026, an Asia hub opened in Singapore, and a Japan data residency programme was announced in July 2026.
What it does
- Generate diagrams, tables and summaries from the content already sitting on the canvas
- Run AI agents, called Sidekicks, that build directly on the board
- Build and execute multi-step visual AI workflows with Flows, then share them across the organisation
- Connect external AI tools through the MCP server: Claude, Cursor, GitHub Copilot and Gemini CLI
- Pull live data from Slack, Jira or Granola onto the board with Connectors, and sync two ways with Jira, Asana, Linear, ClickUp and Azure
- Cluster sticky notes automatically and turn workshop answers, polls, 2x2 matrices and scales into structured results
- Map processes and technical systems with Mermaid, UML, AWS and BPMN shapes, and turn sketches into clickable prototypes
When to use Miro / When not to
A quick filter to help you decide if Miro is the right fit.
When to use Miro
- Product, design and engineering teams that need one shared space for discovery, roadmapping, process mapping and technical documentation
- Distributed and hybrid teams running remote workshops, brainstorms and retrospectives, with timers, voting and asynchronous TalkTrack narration
- Consultants and agencies that facilitate client workshops and want to reuse repeatable formats through templates and blueprints
- Enterprises in regulated sectors such as financial services, manufacturing and life sciences that need data residency, SSO, SCIM and audit logs
- Teams already working with Claude, Cursor, GitHub Copilot or Gemini CLI that want to plug those tools into a shared canvas through the MCP server
When not to use Miro
- Anyone under 16: the Terms of Service state the service is not intended for, and may not be used by, people below that age
- Large open audiences on the Free plan, which caps editing at three boards and has no Guest or Visitor role, so every person invited becomes a Member and takes a seat
- Organisations that need EU, US, Australian or Japanese data residency and granular AI governance without an Enterprise budget, since both sit behind that plan
- Teams looking for an image generator, a coding IDE or a code management tool: Miro is a generalist visual collaboration platform
- Large headcounts on a fixed budget, as billing is per seat, cost grows linearly with the team and invoice billing is no longer open to new customers in every country
How to use Miro
A typical end-to-end flow, from setup to results.
- Sign up from the homepage with Start for free; no credit card is needed
- Work straight from the browser, or install the apps for iOS, Android, macOS on Apple silicon and Intel, Windows 64-bit, the Microsoft Store and interactive displays
- Open a first board from one of the 6,000+ Miro and Miroverse community templates
- Structure the work into boards and project spaces, add workspaces on Business, and save recurring set-ups as reusable blueprints
- Invite the team with the three roles in mind: Member on a paid seat, Guest on a free account invited to specific boards, and Visitor through a public link with no account
- Trigger AI on the canvas: cluster sticky notes, generate diagrams and summaries, or start a Sidekick from a prebuilt library or a customised one
- Build a visual AI workflow with Flows, turn it into a template and share it across the organisation
- Connect Claude, Cursor, GitHub Copilot or Gemini CLI through the MCP server, available on every plan with a daily call cap that starts at 500 per day on Starter
- Start the 14-day free Business trial when the team needs unlimited workspaces, interactive prototypes or the full shape library
- Administer from the Admin Console, where billing and invoices live and where Enterprise adds SCIM provisioning through Okta or Entra ID, with Miro Academy, the Help Center, webinars, Solution Partners and Professional Services to support the rollout
Pros & Cons
Pros
- A permanent free plan with no credit card, including 10 AI credits a month, alongside a self-service 14-day Business trial
- Unusually broad functional coverage in one place: whiteboarding, diagrams, docs, tables, slides, kanban, prototypes and workshop facilitation
- Openness to third-party AI, with the MCP server available on every plan and 250+ integrations offering two-way sync with work management tools
- A documented security posture: ISO 42001, ISO 27001, SOC 2, TISAX, NIST, GDPR compliance, a published DPA and a public subprocessor list
- Data residency across four zones, EU, US, Australia and Japan, with TLS 1.2 or above in transit and AES-256 at rest on every plan at no extra cost
- An explicit statement that customer data is not used to train Miro's AI models
- Access-widening pricing: 20% off with annual billing, free for accredited education, discounts for nonprofits and credits for startups, backed by Miro Academy, the Help Center, community, Solution Partners and Professional Services
Cons
- The Free plan allows only three editable boards and has no Guest or Visitor role, so everyone invited becomes a Member and consumes a paid seat
- AI credits are capped by plan, pooled across the organisation and never carried over from one month to the next; on Free the remaining balance is not even displayed
- AI governance, meaning granular access control, automatic blocking on sensitive boards, attribution logs and the usage dashboard, is Enterprise-only, as is data residency, and Enterprise pricing is not published
- The advertised 8 and 20 USD rates assume annual billing; monthly costs 20% more and that amount is not published on the pricing page
- There is a gap between the commercial promise that data is never used to train Miro AI models and AI Terms that provide for prior consent on Enterprise and a means of objection on the other plans
- Miro warrants neither the accuracy of Outputs nor their non-infringement, and intellectual property claims linked to Outputs are excluded from its indemnification
- Practical friction adds up: no support email is published, invoice billing carries licence minimums of ten on Starter and five on Business and is closed to new customers in some countries, seat reductions only apply at renewal, Enterprise Guard is a paid add-on, and the integrations page linked in the footer returned a 404 when the record was collected
Pricing & Plans
Miro follows a freemium model. A permanent free plan is available at no cost, with three editable boards, unlimited members and 10 AI credits per month. The entry paid tier, Starter, is priced at USD 8 per member per month, and Business at USD 20 per member per month; both figures assume annual billing, monthly billing being 20% more expensive without its amount being published. Enterprise pricing is quoted on request. Billing is per seat, with seats added mid-cycle charged pro rata and reductions taking effect only at renewal. Additional AI credit packages can be purchased on Starter, Business and Enterprise, and Enterprise Guard is a separate paid module. Payment is by card, monthly or annual with no minimum, or by invoice, annual only with a minimum of ten Starter or five Business licences, settled by card, US ACH or wire transfer and processed through Stripe. Subscriptions renew automatically and must be cancelled before the renewal date.
- three editable boards
- unlimited members and 10 AI credits per month
- unlimited private boards
- visitor editing
- high-resolution exports
- unlimited project spaces and blueprints
- timer and voting
- 25 AI credits per member per month
- 500 MCP calls per day
- custom fonts and styles
- unlimited workspaces
- unlimited guests
- 4
- 700+ shapes covering Mermaid
- UML
- AWS and BPMN
- data tables
- Engage activities
- licensing at scale
- enterprise security
- domain control
- SCIM
- audit logs
- EU
- US
- Australia and Japan data residency
- Education free for accredited institutions
- a nonprofit discount
- startup credits
- a 25% Enterprise discount for venture-backed companies
- and a Consultants and Agencies track
Data, GDPR & hosting
A consolidated view of how Miro handles your data.
GDPR overview
Compliance is documented rather than merely claimed. A GDPR badge in the footer links to Miro's trust portal, a Customer Data Processing Addendum is published online and the subprocessor list is available as a public document. The declared controller is RealtimeBoard, Inc. dba Miro, ATTN: Privacy Team, in San Francisco; the Article 27 representative in the EU is RealtimeBoard BV in Amsterdam. Transfers outside the EEA rely on the European Commission's standard contractual clauses under Article 46(2)(c) and on certification to the EU-US Data Privacy Framework, its UK extension and the Swiss-US DPF. The privacy policy lists rights of access, rectification, erasure, restriction, portability, objection and withdrawal of consent, and commits to cooperating with the competent supervisory authority, the UK ICO and the Swiss FDPIC, with binding arbitration as a last resort and FTC investigatory authority acknowledged. Privacy requests go to privacy@miro.com or through the Help Center form. Minimum age: 16.
Who owns the data?
The Terms of Service state that, between the parties, the customer retains all intellectual property and other rights in Customer Content and Customer Materials provided to Miro. The AI Terms extend this to generative features: you retain your ownership rights in Input and you own the Output, with Miro assigning any rights it may hold in that Output. Two reservations apply. Outputs are not exclusive, and Miro warns they may be similar or identical to those provided to other customers. Where a third party powers an AI feature, Miro only undertakes to inform you of any restriction on Output ownership. Deleted content may also survive in routine backups, under the same confidentiality obligations.
Reuse rights
What you create in Miro stays yours to reuse without asking permission: since Input and Output remain the customer's, boards, diagrams, documents and AI-generated content can be exported, published or fed into your own processes freely. The limits are contractual rather than technical, the AI Terms prohibiting the use of Outputs to train competing AI models or to drive automated decisions without human review. In the other direction, Miro states twice on its own pages that your data is not used to train Miro AI models. The AI Terms of 17 July 2024 are more nuanced: under an enterprise Master Cloud Agreement, Miro seeks prior consent, withdrawable at any time, before using Input and Output to improve its AI features, including training and fine-tuning; on free, starter and business plans it undertakes instead to provide "reasonable means to object to such use". Where consent is given or no objection is raised, the licence granted is non-exclusive, perpetual, irrevocable, worldwide and royalty-free, with direct identifiers removed. Personal data may be transferred to the United States under standard contractual clauses and the Data Privacy Framework.
Data retention & training
Hosting summary
Miro offers data residency in four zones, the European Union, the United States, Australia and Japan, as an Enterprise plan option; the Japan residency programme was announced on 8 July 2026. Outside that option, the privacy policy states that data may be transferred to and processed in the United States and in other countries. Those transfers rely on the European Commission's standard contractual clauses and on Data Privacy Framework certification, covering the EU-US framework, its UK extension and the Swiss-US framework. Encryption is TLS 1.2 or above in transit and AES-256 at rest, on every plan and at no extra cost. The subprocessor list is published, and the Help Center documents where AI processing takes place. One caution when reading technical signals: the IP address the domain resolves to (143.204.55.17, Zurich, AS16509 Amazon) is a CloudFront edge node. It shows how the website is delivered, not where customer data is stored, and the two should not be confused.
Things to keep in mind
Risks and trade-offs to weigh before adopting Miro.
- Check your plan against the AI Terms: the promise that data is never used to train Miro AI models coexists with a consent mechanism on Enterprise and a means of objection on the other plans
- The published 8 and 20 USD prices assume annual billing, monthly costs 20% more without its amount being shown, and seat reductions only take effect at renewal, so headcount changes are paid for in advance
- On the Free plan every person invited becomes a Member, since there is no Guest or Visitor role: an open workshop can quietly turn into a bill
- AI credits do not roll over and the balance is invisible on the Free plan, so a team can hit a wall mid-month with no warning
- Miro guarantees neither the accuracy of Outputs nor their non-infringement, warns that Outputs may be similar or identical to those provided to other customers, and excludes such intellectual property claims from its indemnification: generated material still needs human review before it leaves the board
- Using Outputs to train competing models, or for automated decisions without human review, is contractually prohibited, and third-party providers behind some AI features may impose additional terms passed through the subprocessor list
- The entity named on the legal notice is Miro DACH GmbH in Berlin, while the contract is entered into with RealtimeBoard, Inc. dba Miro in San Francisco: worth confirming which entity your procurement and privacy paperwork should name
Setup & Integrations
Technical difficulty
Low for everyday use, higher for an organisation-wide rollout. Signing up is self-service and needs no credit card, the workspace runs in a browser with nothing to install, and desktop and mobile apps are optional. Thousands of templates and Miro Academy shorten the first session, and no technical skill is required to work on a board. Administration is another matter: SCIM provisioning, SSO through Okta or Entra ID, domain control, audit logs and data residency belong to an enterprise deployment, as do the MCP server and the developer platform. Professional Services and a Solution Partner network exist for that stage.
Deployment
Apps stores
Integrations
Supported languages
Behind Miro
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
Does Miro have a free plan?
Can I try a paid plan before committing?
How much does the first paid plan cost?
Does Miro train its AI models on my data?
What are AI credits?
Can Miro connect to my other AI tools?
Where is my data hosted?
Which certifications does Miro hold?
Are there desktop and mobile apps?
How do I contact support?
Should you pick Miro?
Miro has moved from online whiteboard to something more ambitious: a collaboration layer between teams and AI agents. The differentiator it claims is context. Individual AI tools only know a prompt, whereas a shared canvas holds everything a team has already built, and the MCP server pushes that context out to Claude, Cursor, GitHub Copilot or Gemini CLI. The recent trajectory supports the claim: the MCP server launched in February 2026 with Anthropic, AWS, GitHub, Google and Windsurf, Reforge was acquired in March 2026, an Asia hub opened in Singapore and a Japan data residency programme followed in July 2026.
Enterprise maturity is substantive rather than declarative: ISO 42001 alongside ISO 27001, SOC 2, TISAX and NIST, a published processing addendum, a public subprocessor list, residency across four zones and a set of AI governance controls. The trade-off is where those controls sit. Granular AI access rules, automatic blocking on sensitive boards, attribution logs and residency are all Enterprise features, and Enterprise pricing is not published.
Cost is predictable but linear: USD 8 and USD 20 per member per month on annual billing, every seat counted, and on the Free plan every invited person becoming a Member. Two points deserve a buyer's attention. AI credits are pooled and never carried over, so heavy AI use is metered rather than flat. And there is a gap between the marketing line that data is never used to train Miro's models and AI Terms describing consent on Enterprise and a means of objection elsewhere, worth checking against the plan actually signed.
For organisations already working across many tools, Miro is a credible shared surface. For small teams on a tight budget, the free tier is a genuine but narrow entry point.
- Choosing a selection results in a full page refresh.
- Opens in a new window.