Two new image models from Higgsfield are live in the Create Image tool today.
Soul 2.0 is the next generation of Higgsfield's flagship people-and-fashion engine. Soul Cinema is a new model built specifically for cinematic film-still output. They sit alongside every other model in the tool and work with your Visual DNA characters, aspect ratio choices, and prompts exactly as you would expect.
<video src="https://kolboai-development.ams3.digitaloceanspaces.com/dashboard-media/carousel/higgsfield-soul/higgsfield-soul-outro-social.mp4" poster="https://kolboai-development.ams3.digitaloceanspaces.com/dashboard-media/carousel/higgsfield-soul/higgsfield-soul-16x9-v2.png" controls playsinline style="width:100%;border-radius:10px;display:block;margin-bottom:8px;"
Your browser does not support the video tag.
Higgsfield Soul 2.0
Soul 2.0 is built for people. Where most image models treat a human subject as one element among many, Soul 2.0 is engineered around the face, the body, the fabric, and the light falling across them. The results read like editorial photography: sharp, intentional, fashion-grade.
What it does well:
- Magazine covers and fashion editorials - the model understands how a photographer frames a subject
- Character portraits where likeness and expression matter
- Clothing and texture rendering at a level most general models miss
- High-contrast, stylized looks that still read as photographic
Style presets: 33 curated presets ship with Soul 2.0. Each one is a named visual direction - a lighting mood, a color treatment, a compositional energy - so you can dial in the look before you write a word of prompt. Combine a preset with your own text and the model layers them together rather than picking one or the other.
Pricing: from 2 credits per image.
Higgsfield Soul Cinema
Soul Cinema does one thing: it makes images that look like frames pulled from a real film.
Not "cinematic style." Actual movie-frame rendering - the grain structure of a film stock, the color grade of a specific era of cinema, the way a wide lens compresses a background or a telephoto stacks the layers of a scene. The lighting is directional and motivated, not the even, soft-box quality of most AI image output.
What makes it distinct:
- Film-stock grain that reads as photochemical, not digital noise
- Motivated practical lighting (the light always has a source)
- Color grades calibrated to specific cinematic traditions - not just "warm" or "cool" but a full color-science treatment
- Depth and foreground-background separation that looks like real glass and real aperture
Best for: film stills, mood boards for video production, scene references, poster concepts, editorial content where photographic realism alone is not enough.
Pricing: 10 credits per image.
Two bonus Souls
They also come with company: Soul Cast turns a single prompt into a complete character casting sheet - front, back and close-up of the same character - and Soul Location builds entire cinematic worlds and environments, no people in frame.
How to use them
Both models are in the Create Image tool now. Select Higgsfield Soul 2.0 or Higgsfield Soul Cinema from the model picker, write your prompt, and generate. Soul 2.0's 33 style presets appear when that model is active. Soul Cinema works from your prompt directly.
If you use Cinema Controls - the Looks, Style, and Camera chips above the canvas - they compose with both models. Soul Cinema plus a cinematography preset is a particularly strong combination.
Both models are available to all Kolbo.AI subscribers.
