Verdapix is an independent service, not affiliated with OpenMOSS

Verdapix
Back to home

Model Disclosure

Which AI model generates images on Verdapix, who hosts it, its license, and its known limitations.

Last updated: 2026-10-21

Which model generates your images

Images are generated with the open-source MOSS-Image model (OpenMOSS, Apache-2.0), hosted on fal.ai as the model endpoint fal-ai/moss-image. No other image model is used.

  • Model: MOSS-Image, a text-to-image model released by the OpenMOSS team under the Apache-2.0 license.
  • Host: fal.ai runs the model for us. Your prompt and the requested size are sent to fal.ai to generate the image.
  • No training on your data: we do not use your prompts or images to train models.

Verdapix is an independent website and is not affiliated with, endorsed by, or sponsored by OpenMOSS or fal.ai.

Model provenance

  • Model: MOSS-Image
  • Publisher: OpenMOSS
  • Source code: github.com/OpenMOSS/MOSS-Image
  • Model weights: OpenMOSS-Team/MOSS-Image-Base and OpenMOSS-Team/MOSS-Image-SFT on Hugging Face
  • License: Apache-2.0
  • Architecture: 5B-parameter diffusion transformer (DiT), text-to-image
  • Weights: open weights, publicly downloadable
  • Hosting: fal.ai (fal-ai/moss-image)
  • Safety layers: keyword blocklist and AI text moderation on prompts; fal.ai's safety checker and AI image moderation on every output (details below)

Safety systems

Prompts are checked before generation, and every output is checked before it is displayed:

  1. Keyword blocklist — prompts that match our blocked-content rules are rejected before anything is generated.
  2. AI prompt review — every prompt is also reviewed by SeeAPI's text moderation model (text-nsfw-filter). Prompts it flags are rejected.
  3. Host safety checker — fal.ai's safety checker runs on every generated image; flagged images are discarded.
  4. AI image review — every remaining image is reviewed by SeeAPI's image moderation model (nsfw-filter) before it is stored or shown to you. Flagged images are discarded.

If a moderation service is unavailable, the request fails safely and nothing is shown. Blocked requests are refunded in credits as described in the Refund Policy. Rules for what you may create are in the Acceptable Use Policy.

Limitations

AI images can contain mistakes: misspelled or garbled text, wrong hands or objects, or unintended resemblance to real people, brands or artworks. Review every image before you use it, and do not present AI images as real photographs of events or people.

Labeling

Images are AI-generated. If you publish them where disclosure is required by law or platform rules, label them as AI-generated.

Questions

Email support@mossimagegenerator.com.