A Higgsfield MCP alternative

MCP for media generation
— a Higgsfield alternative

Looking for Higgsfield MCP? Twelver is an MCP server for media generation for Claude, Cursor and any client: images and video first, plus music and voiceover — via top models (Nano Banana, Seedream, Seedance, Veo, Hailuo, MiniMax). Via a connector URL and OAuth sign-in, on one balance across every model.

What you can generate

Images, video, music and voiceover right in your AI agent — without a separate service.

🎨

Images

Text-to-image and reference-based images via Nano Banana and Seedream.

🎬

Video

Text-to-video and photo animation via Seedance, Veo and Hailuo.

🎵

Music and audio

As an extra — a standalone music track or text voiceover via MiniMax and TTS (as separate files).

🔎

Search → video from guides

The agent finds the best guides via web search and assembles a finished clip from them.

Connecting

Check the models you want — connector URLs and config assemble themselves. Add them in Claude or to Cursor’s mcp.json, OAuth sign-in.

No keys: paste the URL into your client and sign in — Claude, Cursor and any other OAuth client.

1

Copy the connector URL

2

Paste it into your MCP client

3

Sign in via OAuth — done

Build your server

Check the tools — URLs and config update themselves.

Connector URLs

Image Generation (Nano Banana)

https://twelver.ru/api/mcp/v1/image-generation

Video Generation (Seedance)

https://twelver.ru/api/mcp/v1/seedance-video-generation

Video Generation (Veo)

https://twelver.ru/api/mcp/v1/google-video-generation

Music (MiniMax)

https://twelver.ru/api/mcp/v1/minimax-music

Speech (Google)

https://twelver.ru/api/mcp/v1/google-audio-speech

Web Search

https://twelver.ru/api/mcp/v1/web-search
{
  "mcpServers": {
    "twelver-image-generation": {
      "url": "https://twelver.ru/api/mcp/v1/image-generation"
    },
    "twelver-seedance-video-generation": {
      "url": "https://twelver.ru/api/mcp/v1/seedance-video-generation"
    },
    "twelver-google-video-generation": {
      "url": "https://twelver.ru/api/mcp/v1/google-video-generation"
    },
    "twelver-minimax-music": {
      "url": "https://twelver.ru/api/mcp/v1/minimax-music"
    },
    "twelver-google-audio-speech": {
      "url": "https://twelver.ru/api/mcp/v1/google-audio-speech"
    },
    "twelver-web-search": {
      "url": "https://twelver.ru/api/mcp/v1/web-search"
    }
  }
}

The same mcp.json works in LM Studio, Windsurf, Cline, VS Code and any client that reads an mcp.json.

Try it without signing up. Register and we’ll add more free tokens — plus extra for a few simple onboarding tasks.

How it differs from Higgsfield

Honestly: this is not Higgsfield, but an alternative via top models.

🧠

Top models

Nano Banana and Seedream for images, Seedance / Veo / Hailuo for video, MiniMax for music — pick per task.

💸

Priced in rubles, noticeably lower

Pay by card in rubles, as you go — no currency conversion and noticeably cheaper.

🧩

More than media

The same balance unlocks web search and voiceover — chains like “find a guide → make a video” work out of the box.

🔌

Any MCP client

Claude, Cursor and others — a remote HTTP server behind a URL, OAuth sign-in, no local install.

FAQ

No. Twelver is not affiliated with Higgsfield and does not use its models. It’s a standalone MCP server for media generation via top models (Nano Banana, Seedance, Veo and more) — an alternative for anyone who wants a media generator over MCP.
For images — Nano Banana and Seedream; for video — Seedance, Veo and Hailuo; for music — MiniMax; plus voiceover (TTS). The specific model is chosen by the server URL you connect.
Priced in rubles, noticeably lower. And it is not only media: the same balance covers web search and voiceover, so the agent can find guides and assemble a finished clip from them.
First requests are free after sign-up, then pay-as-you-go from a single balance. Works in any MCP client: Claude Desktop, Claude Code, Cursor and others.

Earn free tokens

Complete onboarding goals to top up your free balance.

  • Create an account

    +150,000

    Sign up to save your chats, agents, and preferences across devices.

  • Run your first web search

    +20,000

    Ask a question that needs fresh information — the AI will search the web and cite its sources.

  • Create your first document

    +20,000

    Generate a structured document — article, report, plan — right from chat and keep editing it collaboratively.

  • Upload your first image

    +20,000

    Drop an image into the chat and the AI can see it — ask questions about its contents or request edits.

  • Generate your first audio clip

    +20,000

    Generate speech, a voice-over, or a sound effect from a text prompt.

  • Generate your first video

    +20,000

    Pick a video-capable agent (or create one tailored for video), then describe the scene you want to bring to life.

  • Generate your first image

    +20,000

    Ask the chat to generate an image — try different styles, aspect ratios, and detailed prompts for best results.

  • Create your first agent

    +50,000

    Design a custom AI agent with its own persona, tools, and instructions.

  • Share something you made

    +20,000

    Send a track, picture or clip you made to someone. They get a page with your work on it — no account needed to look.

  • Top up your balance

    +100,000

    Top up your balance to unlock premium models and tools. Tokens stay available for as long as your subscription is active — they expire when the subscription ends.

  • Top up your balance twice

    +100,000

    Your second paid top-up unlocks an extra bonus on top of the tokens you purchased.

  • Bring a friend who stays

    +200,000

    When someone who arrived through your link makes their first purchase, you get tokens.

  • Bring three friends who stay

    +500,000

    Three friends who stayed. This one is the milestone, and it pays like one.

Image, video and music generation over MCP

A Higgsfield MCP alternative via top models, on one balance — for Claude, Cursor and any client.