Generate with MOSS-Image in your browser
The online studio sends your prompt and selected canvas settings to the hosted fal.ai MOSS-Image SFT endpoint. It returns one image at a time, shows the seed, and lets you open or download the result.
You do not need to download model weights or keep a GPU running. Email verification creates your account and protects its image credits.
Open the online studioStart with three images and no credit card
A verified account receives three one-time starter image credits. No subscription begins automatically, and no payment details are requested for those starter images. Generation remains subject to the shared daily service budget and cooldown.
Use text prompts or structured JSON layouts
Text mode is the quickest starting point for illustrations, product concepts, posters, and editorial scenes. Quote any wording that should appear inside the image, then proofread the result before publishing it.
JSON Layout mode accepts a scene summary plus intended regions for text and objects. The coordinates guide composition but do not guarantee pixel-perfect placement.
Choose online convenience or local control
| Path | What you manage | Best for |
|---|---|---|
| Online studio | Prompt, layout, canvas, and credits | Trying SFT quickly without installation |
| Local ComfyUI | Weights, dependencies, memory, workflow, and hardware | Base access, custom workflows, and self-hosted generation |
Base is available through the local workflow, while this hosted studio uses SFT. If you need checkpoint-level control, continue with the local installation guide or download a matching ComfyUI workflow.