1. Conduid
  2. Content
  3. Higgsfield
MCP server · Content

Higgsfield

· github.com/s0uvik/higgsfield-mcp

MCP server for Higgsfield AI cinematic image and video generation

Unclaimed mcphiggsfieldaiimage-generationvideo-generationtalking-headclaudemodel-context-protocol
InstallView source
37Low

Scored 2 months ago · breakdown

About Higgsfield

Higgsfield is an MCP server in the Content category: mCP server for Higgsfield AI cinematic image and video generation. It has been installed 0 times through Conduid.

Install

Claude Code
claude mcp add higgsfield -- npx -y higgsfield-mcp
npx
npx -y higgsfield-mcp

This server has no ConduID identity, so agent calls to it are not receipted. Pin the version you install and review the source before granting it credentials.

Ask AI

✦

Ask AI about Higgsfield

Powered by Claude · Grounded in docs

I know everything about Higgsfield. Ask me about installation, configuration, usage, or troubleshooting.

Security checks

  • ·README presentNot checked yet.
  • ·License declaredNot checked yet.
  • ·Tests presentNot checked yet.
  • ·Dependencies pinnedNot checked yet.
  • ·No dynamic code executionNot checked yet.
  • ·Scoped permissionsNot checked yet.

README

Higgsfield MCP

MCP server for Higgsfield AI — generate images and videos directly from Claude or any MCP-compatible client.

Setup

Node.js 18+ required. Get your API keys at https://cloud.higgsfield.ai/api-keys.

Claude Desktop

Edit %APPDATA%\Claude\claude_desktop_config.json (Windows) or ~/Library/Application Support/Claude/claude_desktop_config.json (macOS):

{
  "mcpServers": {
    "higgsfield": {
      "command": "npx",
      "args": ["-y", "higgsfield-mcp"],
      "env": {
        "HF_API_KEY": "your_api_key",
        "HF_SECRET": "your_secret"
      }
    }
  }
}

Claude Code

claude mcp add higgsfield -- npx -y higgsfield-mcp
export HF_API_KEY=your_api_key
export HF_SECRET=your_secret

3. Get Your API Keys

  1. Go to https://cloud.higgsfield.ai/api-keys
  2. Create or copy your API Key and Secret

4. Connect to an MCP Client

Claude Desktop

Edit your Claude Desktop config:

  • macOS — ~/Library/Application Support/Claude/claude_desktop_config.json
  • Windows — %APPDATA%\Claude\claude_desktop_config.json
{
  "mcpServers": {
    "higgsfield": {
      "command": "npx",
      "args": ["-y", "higgsfield-mcp"],
      "env": {
        "HF_API_KEY": "hf_xxxxxxxxxxxx",
        "HF_SECRET": "sk_xxxxxxxxxxxx"
      }
    }
  }
}

Restart Claude Desktop after saving.

Claude Code (CLI)

claude mcp add higgsfield -- npx -y higgsfield-mcp

Then set your environment variables before running Claude Code:

export HF_API_KEY=hf_xxxxxxxxxxxx
export HF_SECRET=sk_xxxxxxxxxxxx

Cursor / VS Code

Add to your .cursor/mcp.json or .vscode/mcp.json:

{
  "mcpServers": {
    "higgsfield": {
      "command": "npx",
      "args": ["-y", "higgsfield-mcp"],
      "env": {
        "HF_API_KEY": "hf_xxxxxxxxxxxx",
        "HF_SECRET": "sk_xxxxxxxxxxxx"
      }
    }
  }
}

Tools

Image Generation

Tool Model Notes
generate_image Soul Supports style presets and character references
generate_image_reve Reve
generate_image_seedream Seedream v4 Optional camera_fixed param
edit_image_seedream Seedream v4 Edit Edit existing images via prompt

Video Generation

Tool Model Notes
generate_video DoP (lite/turbo/standard) Requires a motion_id
generate_video_kling Kling v2.1 Pro Prompt = camera movement instructions
generate_video_seedance Seedance v1 Pro Prompt = movement/action description
generate_video_dop_standard DoP Standard Optional duration (2–10s)
generate_talking_head Speak v2 Portrait + WAV audio → lip-synced video

Status & Control

Tool Notes
get_generation_status Poll Soul/DoP/TalkingHead jobs (job_set_id)
get_request_status Poll all other generation jobs (request_id)
cancel_request Cancel a queued job

Characters

Tool Notes
create_character 1–5 face images → reusable character_id. Costs 40 credits.
list_characters All characters with status and thumbnails
get_character Single character by ID
delete_character Permanently delete a character

Upload

Tool Notes
upload_image Upload base64 image → get public URL for use in generation tools

Lookup

Tool Notes
list_styles Style presets for generate_image
list_motions Motion presets for generate_video
debug_credentials Verify keys are loaded

Resources

Browse these before generating:

  • higgsfield://styles — style IDs for generate_image
  • higgsfield://motions — motion IDs for generate_video
  • higgsfield://characters — your character references

How Jobs Work

Tools return either a job_set_id or request_id. Poll the matching status tool until completed, then download the output URL. Results are kept for 7 days.

generate_image(...)             → job_set_id  → get_generation_status(job_set_id)
generate_image_reve(...)        → request_id  → get_request_status(request_id)
generate_video_kling(...)       → request_id  → get_request_status(request_id)

All tools accept an optional webhook_url to receive a callback when the job finishes.

Talking Head Note

Audio must be WAV format. Convert with:

ffmpeg -i speech.mp3 -acodec pcm_s16le -ar 44100 speech.wav

Credits

Operation Credits USD
Image 720p 1.5 $0.09
Image 1080p 3 $0.19
Video lite 2 $0.13
Video turbo 6.5 $0.41
Video standard 9 $0.56
Character (one-time) 40 $2.50

Top up at https://cloud.higgsfield.ai/credits

Links

README mirrored from the source repository 2 months ago. The original is authoritative.

Questions

About Higgsfield

How do I install Higgsfield?

Run claude mcp add higgsfield -- npx -y higgsfield-mcp, then add the server to your MCP client's configuration. Conduid has recorded 0 installs, so the command is known to work with current clients.

Is Higgsfield safe to use with an AI agent?

Its trust score is 37 out of 100 (low). Conduid hasn't run static security checks on this repository yet, so review the source yourself before granting it credentials. It has no ConduID identity yet, so agent calls to it are not receipted.

Is Higgsfield still maintained?

Conduid hasn't recorded a commit date for this repository yet. Check the repository directly for recent activity.