The best AI models for character design
One character who looks the same in every frame. These are Call Sheet's picks for this job, in order, with the reason for each. Call Sheet is a free tool by Melissa Beery that ranks 101 image, video and sound models for a creative job. Its data was last updated on 20 October 2026.
1
Stable Diffusion / ComfyUI
Stability + open community
Trains on your character, so it stays on model across a whole show.
Maximum control.
- Cost
- Free / local (your GPU). High setup effort
- Where
- OpenArt
- Rights
- You own the pipeline and can train on cleared data, so you control where the data goes.
- Watch out
- It runs on your own machine, so setup is steep.
How to ask itpositive: [subject, style, detail tags] | negative: [unwanted] | + ControlNet / LoRA / IP-Adapter conditioningSplit it into a positive prompt and a negative prompt.
Keep the positive prompt on subject, style and detail. Use the negative prompt only to remove artifacts.
Open Stable Diffusion / ComfyUI in Call Sheet →
2
Midjourney V8.2
Midjourney
Its character and style references keep one character and one look across the boards.
Unmatched art direction through style references.
- Cost
- Own platform / subscription
- Where
- Midjourney's own site · check in app: KREA
- Rights
- Web‑trained.
- Watch out
- Won't follow a brief literally.
How to ask it[subject] [doing what], [medium/style], [lighting], [composition] --ar 16:9 --stylize <0-1000>Keep it under about forty words. This model rewards concision and ignores over‑specification.
Order matters: subject, then medium and style, then lighting, then composition.
Open Midjourney V8.2 in Call Sheet →
3
FLUX.2 (pro / flex / max)
Black Forest Labs
Multi‑reference editing keeps design details locked.
Literal prompt adherence and photoreal control.
- Cost
- Mid. Local open variant available
- Where
- Magnific, Higgsfield, Runway · check in app: KREA, OpenArt
- Rights
- Check terms.
- Watch out
- Less painterly than Midjourney.
How to ask it[Subject] + [action] + [style] + [context] + [lighting] + [technical]Describe only what should be present. There is no negative prompt, so "no cars" can put cars in the frame.
Be literal and specific. This model follows exactly what is written, and word order matters.
Open FLUX.2 (pro / flex / max) in Call Sheet →
4
Nano Banana Pro / Imagen (Google)
Google
Up to 14 reference images to composite a character into new shots.
Precision + real‑world knowledge.
- Cost
- Mid (Pro). A faster/cheaper tier exists
- Where
- Magnific, Higgsfield, KREA, OpenArt, Runway
- Rights
- Google terms, plus a SynthID watermark.
- Watch out
- Less 'art‑directed' personality than Midjourney out of the box.
How to ask it[Subject] + [action] + [location] + [composition/shot] + [style & lighting]. Edits: [what to change] + [what to keep] + [refs]Write full descriptive sentences, not keyword lists.
Put any on‑image text inside double quotes, spelled exactly as it should appear.
Open Nano Banana Pro / Imagen (Google) in Call Sheet →