Higgsfield MCP
Connect to Higgsfield via MCP to generate images and videos and manage your creation history
About Higgsfield MCP
Higgsfield MCP lets MCP-capable AI Agents call Higgsfield in a single session to generate images and videos, train consistent characters, and browse creation history. It offers 30+ models, which the agent can select automatically or you can specify manually; it supports text prompts, reference images, or a combination of both, with images up to 4K, videos up to 15 seconds, and a range of cinematic styles. To connect, just add the server URL and authenticate with your Higgsfield account — no API key management required.
Best for
Best use cases
Generate images in an MCP-capable agent
Create videos in an MCP-capable agent
Generate content from reference images
Train and maintain character consistency
Features
Multi-model generation
Offers 30+ models; the agent can automatically pick the best fit, or you can specify one manually.
Image and video creation
Supports text prompts, reference images, or combined input; images up to 4K, videos up to 15 seconds.
Character training
Consistent character generation via Soul training.
MCP agent connection
Works with MCP-capable agents and authenticates through a Higgsfield account, with no API keys to manage.
Pros
- Supports 30+ models, selected automatically or specified manually
- Supports text prompts, reference images, or combined input
- No API keys to manage or configure
- Generate images, create videos, train characters, and browse history in the same session
Cons
- Videos are limited to 15 seconds
Pricing Plans
Free trial
- 100 credits
* Pricing information is for reference only. Please refer to the latest announcement on Higgsfield MCP's official website.