Sign in Get started

Havincy MCP & API

Havincy MCP for AI agents

Connect Havincy to ChatGPT, Claude or Cursor to generate images, videos, voice-overs, music and viral effects — on brand, with your credits — without leaving the conversation.

1

Add the Havincy connector

In Settings → Apps & Connectors, turn on Developer mode, click Create and paste this URL with OAuth authentication.

https://havincy.com/mcp
2

Start creating in ChatGPT

Sign in to Havincy, allow access, then pick Havincy from the tools menu and just ask.

“Use Havincy to create 4 product shots of my sneaker on a marble table, 1:1.”

1

Add the custom connector

In Settings → Connectors, click Add custom connector, name it Havincy and paste this URL.

https://havincy.com/mcp
2

Start creating in Claude

Click Connect, sign in to Havincy, choose your workspace and permissions, then ask Claude.

“Use Havincy to create 4 product shots of my sneaker on a marble table, 1:1.”

1

Add the server to mcp.json

Open Cursor Settings → MCP, add a server and paste this configuration.

{
  "mcpServers": {
    "havincy": {
      "url": "https://havincy.com/mcp"
    }
  }
}
2

Create from your editor

Cursor opens the Havincy sign-in on first use. Ask for visuals for your landing page, game or app.

“Use Havincy to create 4 product shots of my sneaker on a marble table, 1:1.”

1

Run one command

In your terminal, add Havincy as a remote MCP server.

claude mcp add --transport http havincy https://havincy.com/mcp
2

Sign in with /mcp

Type /mcp in Claude Code, select Havincy and sign in. Assets come back straight into your project.

“Use Havincy to create 4 product shots of my sneaker on a marble table, 1:1.”

1

Create an API key

Create a key in Profile → API & MCP, then send it in the Authorization header.

curl https://havincy.com/api/v1/account \
  -H "Authorization: Bearer $HAVINCY_API_KEY"
2

Call the API from your tools

Scripts, n8n, Make, your backend: the same operations as the MCP tools, documented with OpenAPI.

Get an API key

Documentation OpenAPI

Create with Havincy, from your agent

On brand, every time

Your agent reads your brand kits — tone, audience, colors, instructions — and writes prompts that stay true to your brand. Reuse your product photos and characters as references.

Create brand visuals

Your agent picks the model, checks the cost and follows the job

Describe the result, not the settings. The agent lists the models, picks a compatible one or the one you name, checks your credits and waits until the video is ready.

Create a campaign video

Viral video effects from a single photo

Melt, inflate, giant in the city, clones… Send a photo, choose an Encore effect, and get a 9:16 video ready to post with its share link.

Try an Encore effect

Voice-overs and music in the same conversation

Write the script with your agent, then generate the voice-over in the right language and voice — or a soundtrack — without opening another tool.

Create a voice-over

From idea to published post

Once the video is ready, ask your agent to post it or schedule it on TikTok, Instagram Reels or YouTube Shorts through Havincy Diffusion.

Publish from the chat

The latest models, in one place

Give your agent access to the Havincy catalog: video, image, voice, music and 3D models from the leading labs.

Video models
  • Kling 3.0 Pro
  • Seedance 2.5
  • FLUX 3 Vidéo
  • MiniMax H3
  • Wan 3.0
  • PixVerse V6
  • Luma Ray 3.2
Image models
  • Nano Banana 2
  • FLUX Schnell
  • Seedream 5.0 Pro
  • Ideogram V3
  • Qwen Image
  • GPT Image 2.5 Flare
  • Grok Imagine Image 2.0
Audio & 3D models
  • ElevenLabs v4
  • MiniMax Speech 2.8 HD
  • Qwen 3 TTS 1.7B
  • Grok TTS
  • Stable Audio
  • Meshy 6
  • Hunyuan 3D 3.1 Pro

See all models

Why connect Havincy to your agent

No API key, simple setup

Paste one URL and sign in with your Havincy account in the browser. You choose the workspace and permissions.

Your brand built in

Brand kits, characters and your media library are available to the agent, so results stay consistent.

One connection, every model

Kling, Veo, Seedance, GPT Image, Nano Banana, ElevenLabs and more, without managing a provider account for each.

Your credits, full control

Same prices as the app. The agent checks your balance, and you can disconnect an app or revoke a key at any time.

Create, then publish

Generate, refine and post to your social networks from the same conversation — publishing is a separate permission.

Viral effects included

Every Encore effect is available as a tool: one photo in, one vertical video out.

Encore effects, one sentence away

Ask your agent for an effect by name: it sends the photo and returns the video.

Floating Fall You fall in slow motion as your things float around you. Selfception An endless zoom into a mini you, holding a mini you. Wild Ride The camera whips around you in a high-speed orbit. Infinite Clones Your clones keep pouring out of a tiny car. Street Colossus You become a giant towering over the city. Confetti Your photo bursts into a shower of confetti. Earth Zoom The camera flies up from your photo all the way to space. Zombie Your photo turns into a zombie, just in time for Halloween. Cake Surprise: your photo was a cake all along, and it gets sliced. Squish A hand squishes your photo like a stress ball. Inflate Your photo puffs up like a balloon and floats away. Melt Your photo melts like warm wax.

For developers

A REST API for your product and automations

Every MCP tool is also an HTTP endpoint. Integrate Havincy generation into your SaaS, your e-commerce back office or your no-code workflows.

  • API keys with permissions and expiration, per workspace
  • Images, videos, voice, music, 3D and Encore effects
  • OpenAPI spec, rate limits per plan, usage logs
Read the documentation Get an API key
POST /api/v1/generations
curl -X POST https://havincy.com/api/v1/generations \
  -H "Authorization: Bearer $HAVINCY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"modality":"image","model_id":"fal-nano-banana-2","prompt":"A cat astronaut, watercolor","wait_seconds":40}'

Frequently asked questions

How does Havincy connect to AI agents?

Through MCP, the open standard used by ChatGPT, Claude, Cursor and Claude Code. You add the Havincy server URL as a connector, sign in, and your agent can call Havincy's tools in natural language.

Which assistants are supported?

ChatGPT, Claude (web and desktop), Claude Code, Cursor and any client that supports remote MCP servers over HTTP. Clients without OAuth can use an API key.

Do I need an API key?

No for ChatGPT, Claude, Cursor and Claude Code: you sign in with your Havincy account in the browser. API keys are for the REST API and clients that cannot sign in.

How does pricing work?

Generations use your workspace credits at the same price as in the app. Your agent can check your balance, and every call appears in your usage history.

What can I create?

Images (up to 4 at once), videos from text or an image, voice-overs, music, 3D models and Encore effect videos. Available formats, durations and resolutions depend on the chosen model.

Can I choose the model myself?

Yes. Name the model you want, or ask your agent to compare the options for your brief and budget.

Can my agent publish to my social networks?

Yes, if you grant the Publish permission and your plan includes Diffusion. The agent can post now or schedule on TikTok, Instagram Reels or YouTube Shorts.

Is my data safe?

Each connection is limited to the workspace and permissions you choose. You can see connected apps, disconnect them or revoke keys at any time in Profile → API & MCP.

Connect once.
Create everywhere.

Bring Havincy into ChatGPT, Claude and Cursor to generate, refine and publish your content from the conversation.

Connect Havincy to your agent Documentation