# Makeform

> Makeform is a free AI-native form builder that turns a plain-language description into a working form.

Use Makeform to create, edit, publish, embed, and automate forms, surveys, quizzes, and data-collection workflows. The free plan includes unlimited forms and submissions within fair-use guidelines.

Agents should start with this index, load the smallest relevant linked document, and use the REST API or hosted MCP runtime as the operational contract.

## Start here

- [Create a form with AI](https://www.makeform.ai/create): Start an anonymous AI form draft in the browser; no account is required to begin.
- [AI agent guide](https://www.makeform.ai/agents.md): Choose MCP or REST, configure authentication, and follow supported agent workflows.
- [Developer documentation index](https://www.makeform.ai/docs/llms.txt): Discover the current documentation and REST API reference pages; the index follows the live docs navigation.
- [API reference](https://www.makeform.ai/docs/api-reference/introduction.md): Build integrations with the Makeform REST API.
- [MCP runtime manifest](https://mcp.makeform.ai): Discover the live MCP tool categories, recommended paths, safety guidance, and client configuration.
- [AI information](https://www.makeform.ai/ai-info.md): Read the authoritative capability, access, pricing, and limitation summary for AI systems.

## Product

- [Makeform overview](https://www.makeform.ai/index.md): Understand the product, primary workflows, and current access model.
- [Features](https://www.makeform.ai/features.md): Review form building, logic, design, integrations, AI, collaboration, and workflow capabilities.
- [Pricing](https://www.makeform.ai/pricing.md): Compare the Free, Pro, and Business plans.
- [Templates](https://www.makeform.ai/templates.md): Learn how reusable templates accelerate form creation.
- [Free tools](https://www.makeform.ai/tools.md): Explore task-specific AI form generators and utilities.
- [About Makeform](https://www.makeform.ai/about.md): Learn what Makeform builds and why.

## Navigate by interface

- [Documentation](https://www.makeform.ai/docs/introduction.md): Start with Makeform product and developer concepts.
- [REST API](https://www.makeform.ai/docs/api-reference/introduction.md): Use ordinary authenticated HTTP from server-side Python, Node.js, or other runtimes.
- [Hosted MCP](https://mcp.makeform.ai): Connect coding agents and compatible AI clients through Streamable HTTP.
- [Browser Embed SDK](https://www.makeform.ai/agents.md): Embed public forms in websites without exposing server credentials.
- [Makeform Skills and workflow guidance](https://www.makeform.ai/agents.md): Follow supported creation, editing, publishing, submission, and analysis workflows.

## Navigate by goal

- [Create and publish forms](https://www.makeform.ai/agents.md): Create drafts from prompts, inspect configs, apply checked edits, and publish with confirmation.
- [Embed forms](https://www.makeform.ai/features.md): Use hosted, Standard, Popup, Slider, or Fullpage experiences.
- [Collect and analyze submissions](https://www.makeform.ai/features.md): Retrieve, filter, summarize, chart, export, and analyze response data where supported.

## Machine discovery

- [MCP server](https://mcp.makeform.ai): Use Makeform's canonical Streamable HTTP MCP endpoint.
- [MCP Server Card](https://mcp.makeform.ai/server-card): Read experimental machine-readable MCP identity and connection metadata.
- [AI Catalog](https://www.makeform.ai/.well-known/ai-catalog.json): Discover Makeform's public AI artifacts.

## Notes

- `/create` is the public creation entry. `/signin` is the shared login and signup entry for account actions such as saving, publishing, and managing forms.
- Makeform API keys and OAuth access tokens are secrets. Keep them out of browser code, prompts, repositories, and public logs.
- The MCP Server Card and AI Catalog formats are emerging discovery contracts; the REST API and MCP runtime remain the operational interfaces.
- Use exact paths, methods, schemas, tool names, permissions, and safety guidance from the current linked source rather than inferring them from summaries.