55 MODELS · 11 PROVIDERS · 4 MODALITIES
Generate with every model.Keep every result.
The last ferry of the night. She folded the letter into a square, tucked it inside her coat, and did not look back.
Shot · 03
PROMPT
Ferry deck, night, backlit. She faces away from camera; sea wind lifts one corner of her coat. 35mm, grain.
References · from upstream
Flux 2 Pro2.1s
Gemini 3 ProPicked
GPT Image 23.4s
NovelAI V4.5Base model
LoRA stack
Trigger words
PROMPT
Autumn woodland, a fox asleep in fallen leaves. Shallow depth of field, warm light.
TEXT → IMAGE → IMAGE
Change the outfit, change the style, still the same character.
Write the character from one sentence, then hand that image back as the reference for the next edit. Swap only the jacket, or redraw him as a game sprite. The hard part was never the edit; it is still being him afterwards.
- GPT Image 2
- Gemini 3 Pro
- Flux 2 Pro
- NovelAI V4.5
- +4 more
GeneratingIMAGE → VIDEO
Take the one you picked and let it move.
No rewriting the brief. The selected result becomes the first frame, and the prompt, references and camera notes travel with it — so the clip lives in the same world as the still.
- Seedance 2.0 Fast
- Kling 2.6
- Veo 3.1
- Gemini Omni Flash
Seedance 2.0 Fast
IMAGE → 3D
Anything in the archive becomes a reusable asset.
Every saved image can go to 3D. The source, the prompt and the generated model stay on one record, so next time you reach for it instead of hunting for it.
- Rodin Gen-2.5
- Hunyuan3D v3.1 Pro
- Trellis 2