Models / Face Lock
Keep the same face in every shot with Face Lock
PuLID identity conditioning on the FLUX family — lock one reference face and it holds across every scene, outfit and lighting setup, instead of drifting into a different person shot to shot. See sample output, pricing, and how it compares below.
Try Face Lock — freeSpeed
Runs with FLUX sampling
Cost
2 credits / image
Best for
One identity, every scene
Sample output
ReferenceThe reference. Every shot below is locked to this one face — different scene, different light, different wardrobe, same person.
How it compares
FAQ
What is Face Lock?
PuLID identity conditioning: you give it a reference face, and every generation keeps that face instead of inventing a new person each time.
Does Face Lock cost extra?
It runs as part of the normal FLUX sampling pass at the same 2-credit trained-avatar rate as a plain FLUX.1 dev or NoobAI XL render — no separate Face Lock surcharge. It isn't Boost-eligible, though: the uploaded reference photo never leaves Likna's own GPU, so there's no cloud option to upgrade to.
Which models support Face Lock?
The FLUX family. Z-Image Turbo renders straight from your trained avatar without a separate face-lock pass.
Is it the same as training an avatar?
No — training learns you from your uploaded photos and is the stronger option. Face Lock holds a single reference face at generation time, and the two work together.