Moss Image text-to-image
Describe your subject, style, composition and lighting. Moss Image is the default, text-only model powered by GPT Image 2.5 Flare; it does not accept reference images.
IMAGE WORKSPACE
Moss Image · Image workspace
Recent creations
Sign in and create your first image to start your collection.
FREE GPT IMAGE
Your ideas, brought to life with Moss Image.
Start with a text prompt in Moss Image. The default Moss Image model (moss-image) uses GPT Image 2.5 Flare and accepts text only, with no reference images. For photo editing, switch to a GPT Image or Nano Banana model that supports image inputs. Review the credit cost before generating.
More models, same workspace
Keep creating with the available GPT Image and Nano Banana models. Models that support image-to-image let you upload authorized reference photos within their displayed limits.
Clear credit pricing
Image generation uses purchased credits. The cost depends on your model and settings and is shown before submission. Save your completed images promptly; provider download links may expire.
FAQ
A few things to know
What is Moss Image?
Moss Image is an independent AI image creation service. Its default text-to-image model, moss-image, is powered by GPT Image 2.5 Flare. The workspace also offers GPT Image and Nano Banana models.
How do I start generating?
Sign in, purchase a credit plan, describe your image and review the displayed cost. Moss Image is selected by default for text-to-image generation.
Can I upload a reference photo?
The Moss Image model accepts text prompts only and does not accept reference photos. To edit a photo, choose an available GPT Image or Nano Banana model with image-to-image support, and follow its upload limits.
Is image generation free?
Image generation requires purchased credits. New accounts start with 0 credits. Choose a monthly plan, annual plan or one-time credit pack on the pricing page.
What happens if a generation fails?
Credits reserved for a confirmed failed task are returned. A temporary status-check error is not a confirmed failure; check the saved task before submitting again.
Can I choose other models and advanced settings?
Yes. GPT Image and Nano Banana models remain available, with resolution, quality, aspect ratio and reference support depending on the selected model. The workspace shows the cost before you submit.