# APIANT.ai

> APIANT.ai is a hosted integration and automation platform operated by AI agents. You describe an integration or automation in plain language, to the in-app assistant or to your own AI client over MCP, and the agent connects the apps, builds the automation, tests every scenario before it ships, then keeps it running and reports failures. Automations are stored as data, not as code. APIANT has a canvas that your AI agent can control. Free for everyone to start.

Key facts:

- MCP server: https://mcp.apiant.ai (remote MCP over Streamable HTTP, OAuth sign-in with PKCE in the browser, nothing to install; OAuth scopes `mcp:tools` and `mcp:automation_builder`). Setup is documented for Claude Code.
- MCP Registry: published as `ai.apiant/apiant` (https://registry.modelcontextprotocol.io/v0/servers?search=ai.apiant/apiant). Server card: https://apiant.ai/.well-known/mcp/server-card.json. Agent skill: https://apiant.ai/.well-known/agent-skills/apiant/SKILL.md
- Sign-up: self-serve and free at https://app.apiant.ai/register. The free plan needs no credit card; paid plans are picked in billing after sign-up. See https://apiant.ai/pricing.
- Claude plugin: https://github.com/APIANT/apiant-plugins (`/plugin marketplace add APIANT/apiant-plugins`, then `/plugin install apiant@apiant`); it connects Claude to the same MCP server.
- If an app is not in the catalog, the agent reads the vendor's API documentation, builds the connector and its authentication, and tests each operation against the live API. If the documentation cannot be read, it stops rather than guessing.
- Built-in capabilities: scheduled, polling and webhook triggers; conditions, loops, parallel branches, subroutines, fan-out and fan-in; two-way sync with loop prevention; hosted forms; chat widgets that run on the customer's own Anthropic or OpenAI connection; web service endpoints; human approval steps; automatic retries, alerts, execution history and versions.

## When to use APIANT

- Connect an app that has no ready-made connector: the agent reads the vendor's API documentation, builds the connector and its authentication, and tests each operation against the live API.
- Build a multi-step automation with conditions, loops or parallel branches, and have every scenario tested before it is turned on.
- Keep records matched in two apps in both directions, with loop prevention built in.
- Run an integration on a schedule, when an app changes, or the moment a webhook arrives, with retries, alerts and execution history.

## When it is not the right tool

- The app has no API documentation the agent can read (a private or undocumented API). The agent stops rather than guessing, so it can't build that connector.
- You want code you own running on your own server, without paying a subscription. APIANT stores automations as data and runs them on its hosted platform.

## Product

- [AI agent builder](https://apiant.ai/ai-agent-builder): build AI agents and automations by describing them; the agent tests every scenario before it ships
- [Workflow automation](https://apiant.ai/workflow-automation): triggers, logic, two-way sync, retries and alerts, built from a plain-language description
- [MCP server](https://apiant.ai/mcp): connect an AI client to your apps through one remote MCP server
- [App catalog](https://apiant.ai/apps): every app you can connect, with its triggers and actions; each app has a page at https://apiant.ai/connections/{slug}
- [Pricing](https://apiant.ai/pricing): plans by apps, monthly tasks, assistant credits and history retention
- [Customer stories](https://apiant.ai/stories): how teams use APIANT, told from their own run history
  - [Fitness studio: member reports from Mindbody](https://apiant.ai/stories/fitness-studio-member-reports): A studio owner wanted to see which members were drifting away. He asked his AI client for the Mindbody reports, and had seven working by the next afternoon.
  - [Dental practice: Jotform intake into Cliniko](https://apiant.ai/stories/dental-practice-patient-intake): A dental practice asked for its online bookings and health forms to land in Cliniko, matched to the right patient or not written at all.
  - [Gym: Mindbody clients in ActiveCampaign](https://apiant.ai/stories/gym-mindbody-activecampaign-sync): A gym owner wanted ActiveCampaign to know what Mindbody knows. New clients now arrive hourly through the day, tagged with what they actually do.
  - [Wellness spa: Mindbody to Google Calendar](https://apiant.ai/stories/wellness-spa-calendar-sync): A spa owner wanted Mindbody's bookings and time blocks in Google Calendar, kept in step. Its block sync ran 1,508 times in just over two weeks, all successful.
  - [Training studio: Mindbody sales to webhooks](https://apiant.ai/stories/training-studio-mindbody-sale-webhooks): A training studio wanted every new client's first package sale sent from Mindbody to its own systems. The first live sale went through half an hour after signup.
  - [Outreach team: funding news into HubSpot](https://apiant.ai/stories/growth-outreach-funding-signals): An outreach team wanted new funding rounds from the news logged in HubSpot, ready to check before outreach. Its two scans have written 53 round notes since.
  - [Yoga studio: Mindbody sales into a register](https://apiant.ai/stories/yoga-studio-membership-register): A yoga studio wanted every Mindbody sale in its membership register, sorted by its own rules. It went live on day three and sorted 21 paid sales by the afternoon of day four.
  - [Church: Stripe gifts into DonorPerfect](https://apiant.ai/stories/church-stripe-donorperfect-gifts): A church needed each Stripe gift in DonorPerfect, on the right donor, with its own codes. By the next day, its charges were arriving as gifts on their own.
  - [Fitness studio: Mindbody sign-ups to GoHighLevel and Kisi](https://apiant.ai/stories/fitness-studio-mindbody-gohighlevel-kisi): A fitness studio wanted each new Mindbody membership or recovery pass to start its GoHighLevel follow-up and Kisi door access. Both were built in one day.
  - [Veterans wellness: Mindbody sign-ups into Kit](https://apiant.ai/stories/veterans-wellness-mindbody-kit): A wellness provider for veterans wanted every Mindbody sign-up in Kit, tagged by who they are and where they go. Four automations were running the same day.

## For agents

- [Apps index for language models](https://apiant.ai/apps/llms.txt): the app catalog by category, with links to each app's Markdown page
- [agents.md](https://apiant.ai/agents.md): how an AI agent connects to APIANT (remote MCP, OAuth with PKCE, scopes) and where to read
- [auth.md](https://apiant.ai/auth.md): how an agent signs in: OAuth endpoints, dynamic client registration, PKCE, scopes and token lifetimes
- [APIANT plugins for Claude](https://github.com/APIANT/apiant-plugins): the public plugin marketplace repository
- [ARD manifest](https://apiant.ai/.well-known/ard.json): Agentic Resource Discovery entries, with the MCP server card inline
- [OpenAPI spec](https://apiant.ai/openapi.json): OpenAPI 3.1 description of the public app catalog feed (https://app.apiant.ai/catalog/v1, no authentication) and the MCP server's OAuth scopes

## Markdown versions

Every page is also served as Markdown: request it with `Accept: text/markdown`, or add `.md` to its address (for example https://apiant.ai/apps/finance.md or https://apiant.ai/connections/slack.md). The main ones:

- [Home](https://apiant.ai/index.md)
- [Pricing](https://apiant.ai/pricing.md)
- [AI agent builder](https://apiant.ai/ai-agent-builder.md)
- [Workflow automation](https://apiant.ai/workflow-automation.md)
- [MCP server](https://apiant.ai/mcp.md)
- [About](https://apiant.ai/about.md)

## Languages

The marketing pages (home, pricing, about, MCP server, AI agent builder, workflow automation, support, Slack, request an app, and the customer stories) are also published in Simplified Chinese, Hindi, Spanish, French and German, each with its own Markdown twin. English is the reference; the app catalog, the app and pair pages, solutions, the legal pages and the documentation are English only.

- [Simplified Chinese](https://apiant.ai/zh) (https://apiant.ai/zh/index.md)
- [Hindi](https://apiant.ai/hi) (https://apiant.ai/hi/index.md)
- [Spanish](https://apiant.ai/es) (https://apiant.ai/es/index.md)
- [French](https://apiant.ai/fr) (https://apiant.ai/fr/index.md)
- [German](https://apiant.ai/de) (https://apiant.ai/de/index.md)

A translated page lives at the same path under the language prefix, for example https://apiant.ai/es/pricing (Markdown: https://apiant.ai/es/pricing.md).

## Documentation

- [Docs index for language models](https://apiant.ai/docs/llms.txt): every documentation page with a one-line summary
- [Introduction](https://apiant.ai/docs/introduction)
- [Quickstart: build from Claude Code](https://apiant.ai/docs/quickstart/claude-code)
- [APIANT over MCP](https://apiant.ai/docs/mcp)
- [Build a connector, action or trigger](https://apiant.ai/docs/apps/custom-connectors)
- [Testing an automation](https://apiant.ai/docs/automations/testing)
- [App catalog feed reference](https://apiant.ai/docs/reference/apps): the public JSON feed, also described in https://apiant.ai/openapi.json

## Optional

- [About APIANT](https://apiant.ai/about)
- [Request an app](https://apiant.ai/request-app)
- [Terms](https://apiant.ai/tos)
- [Privacy](https://apiant.ai/privacy)