MOSS Image · words into worlds
MOSS Image AI image generator with words that work
Turn your ideas into posters, product visuals and illustrations with MOSS Image. Clear English and Chinese text, thoughtful layouts, and more control over every detail. Your first image is free.
02Aspect ratio
Auto creates a square image.
- Model
- MOSS Image
- Output
- HD image
- Cost
- 2 credits
GPT Image 2 community references · not MOSS-generated
Sign up free: your first MOSS image is on us.
USE CASES / Community inspiration
One brief, many possibilities.
Explore product visuals, posters, interfaces, illustrations and more through GPT Image 2 community references. Load a full prompt and make it your own with MOSS Image.
GPT Image 2 community references from awesome-gpt-image-2, curated by freestylefly and the credited creators. MIT license. These examples were not generated by MOSS Image. Prompts can be adapted for MOSS Image; results will vary.
PROMPT LIBRARY / Community inspiration
Find your next starting point.
496 full community prompts, organized across 13 categories. Original creators and sources are credited on every example.
GPT Image 2 community references from awesome-gpt-image-2, curated by freestylefly and the credited creators. MIT license. These examples were not generated by MOSS Image. Prompts can be adapted for MOSS Image; results will vary.
03 / About
What is MOSS Image?
MOSS Image is an image creation workspace powered by OpenMOSS’s MOSS-Image model through fal.ai. Describe a scene in natural language, write the exact English or Chinese text you need, or use a JSON layout to guide where words and objects belong.
Natural language, structured layouts
Start with a plain-language idea, a detailed description or a JSON layout. MOSS Image supports structured prompts for more deliberate placement of text, objects and visual elements.
English and Chinese, in the picture
Give your poster a headline, your packaging a label or your infographic a caption. MOSS Image focuses on English and Chinese text rendering alongside coherent visual layouts.
No local setup required
Run MOSS-Image-SFT online from your browser. Write a prompt, choose an aspect ratio and generate without installing local software.
Eight aspect ratios
Create square images for feeds, tall posters and stories, or wide banners and thumbnails. Auto generates a square with MOSS Image.
From realistic to imaginative
Explore realistic product photography, illustrated worlds and anime-inspired scenes. Describe the style, lighting and composition to shape your next visual.
496 community prompts
Explore credited GPT Image 2 community references across 13 categories. Load a full prompt into the MOSS Image editor and adapt it for your own image.
This is an independent workspace using models through fal.ai. It is not operated by or affiliated with OpenMOSS or fal.ai.
04 / How it works
From brief to image in four steps
- STEP 01
Start with an idea
Use MOSS Image text-to-image generation to turn your written idea into a new image.
- STEP 02
Write the brief
Describe the subject, setting, light and any text you want in the picture, or start from one of the ready-made prompts.
- STEP 03
Pick a shape and generate
Choose an aspect ratio and press Generate. Each image costs 2 credits. Your task is saved while the model works.
- STEP 04
Download or iterate
Save the image, or reuse the prompt and refine it for another generation. Everything you make stays in your account History.
05 / Spec sheet
Input, output and price at a glance
- Model
- OpenMOSS MOSS-Image-SFT, served through fal.ai.
- Modes
- Text to image: create a new image from a written prompt.
- Input
- A prompt of up to 6,000 characters: natural language, tags or a JSON layout.
- Aspect ratios
- Auto, 1:1, 4:5, 3:4, 2:3, 9:16, 16:9 and 3:2.
- Output
- One image per run, about 1 megapixel. Downloads preserve the image format returned by the model.
- Price
- 2 credits per MOSS image. Your first image is free after sign up, then as low as $0.10 per image.
- Speed
- Generation time varies with model load and the request. Follow progress in the workspace or History.
- Privacy
- History is restricted to your account. Stored images use unlisted media URLs. Remove completed images from History at any time.
GUIDES / Explore the model
A little knowledge. More possibilities.
Meet Base, SFT and the community model Kromoss, or take MOSS Image into your own ComfyUI workflow.
MOSS Image models & Kromoss
Understand the model family, find the right weights, and choose how you want to create.
Read the guide COMFYUI WORKFLOWMOSS Image ComfyUI workflow
A practical setup guide for native ComfyUI, with a separate path for Kromoss and the optional HF loader.
Read the guide06 / Pricing
MOSS Image pricing, plain and simple
$0.10 per image on the Max plan billed yearly. Every image costs 2 credits. Start with a free image, then pick a plan or a one-time pack.
Credit packs
Pay once, no subscription. Pack credits never expire and are used after subscription credits.
How credits work
- 2 credits = 1 image
- Every MOSS Image generation costs 2 credits. Paid images have no watermark and can be used commercially.
- 1 free image after sign up
- Try MOSS Image without a credit card. Download the original model output.
- Yearly saves up to 37%
- Yearly plans are billed once a year and still get their credits month by month. Unused plan credits expire at the end of each month.
- Credit packs never expire
- One-time packs from $19.90. Pack credits are used after plan credits.
- Failed images are refunded
- If a MOSS Image generation fails or is blocked by the safety filter, its credits go straight back to your balance.
- Cancel anytime
- Cancel your subscription from Manage billing in your account menu. Cancellation takes effect at the end of the paid billing period. Prices in USD.
Start free
Your first MOSS image is on us
Sign up and generate one MOSS image free. No credit card, nothing to install.