higgsfield-soul-id
Train a reusable face identity model for personalized image and video generation
by higgsfield-aiFree
Description
higgsfield-soul-id is an agent skill for training a Soul Character, a reusable identity model of a person's face that Higgsfield uses for identity-faithful image and video generation. You provide five to twenty photos of the face from varied angles and lighting, as local file paths or pre-uploaded image IDs. There are two model variants: the image variant, which is the default, and a cinematic variant for video work. Once trained, the character can be reused across all Soul-powered generations. Training requires a paid plan of at least the Basic tier, and the skill tells you before submitting if the account is on the free plan. Before any command it checks that the higgsfield tool is installed, and it asks you to sign in yourself if the account is not authenticated. Use it only with photos of a person who has agreed to have a model made of their face, and keep the model to uses that person has approved. It is part of the higgsfield-ai/skills repository on GitHub, which has about 1,300 stars. The listing shows it as free, and it has about 181,700 installs. Its security audits show passes from Gen Agent Trust Hub and Socket, with a warning from Snyk.…
Highlights
- Trains a reusable face identity model
- Five to twenty photos from varied angles
- Image and cinematic variants
- Paid plan required for training
- Reusable across Soul generations
Ratings & Reviews
No ratings yet
Be the first to rate higgsfield-soul-id.
People also visit
higgsfield-video-explainer
Build a complete non-photoreal narrated explainer or story video
manim-video
Build reusable Manim explainers for technical concepts, graphs, system diagrams
tasteforge-video
File-driven multimodal image, video, and 3D-asset discovery; taste interviews
higgsfield-soul-id FAQ
How many photos are needed?
Five to twenty photos of the face, from varied angles and lighting.
Does training cost money?
Training requires a paid plan of at least Basic. The skill warns you before submitting on the free plan.
Which variant should I choose?
The image variant is the default. The cinematic variant is for video work.
Who should the model be made of?
The skill is for a person who has agreed to a model of their face, with uses they have approved.
How do I install it?
Run npx skills add https://github.com/higgsfield-ai/skills --skill higgsfield-soul-id.
What are the best higgsfield-soul-id alternatives?
Popular alternatives to higgsfield-soul-id include higgsfield-video-explainer, video-frames and manim-video. See all higgsfield-soul-id alternatives