Synas Image 001 generates images you can reproduce. Synas Detect 001 tells you whether an image was AI-generated or edited, and how sure it is. Both are open-weight, so you can run them yourself.
One makes images. The other checks them. Use either alone, or run them in the same loop.
Synas Image 001
Text-to-image generation and editing with deterministic seeds and style locks, so a result you like is a result you can get again.
Synas Detect 001
Looks for signs that an image was AI-generated or edited, and reports a calibrated score with its confidence, never a bare yes or no.
Create images with Image 001. Record the prompt, seed and style lock so anyone can reproduce the result.
Run any image, yours or someone else's, through Detect 001 and read the score together with its confidence.
Both models are open-weight, so the whole loop can run on your own machines, with images staying where you keep them.
Published weights mean you don't have to take a description on trust.
Self-hosting means the images you generate or check never have to leave your own infrastructure.
Researchers and security teams can test a model's behavior directly, on their own images, instead of relying on a vendor's claims.
A downloaded model doesn't change underneath you, which matters when results have to be reproduced or defended later.
Open-weight and open-source are not the same thing. Our guide explains the difference and what to check before you build on any open model.
Plain-language explanations of the ideas behind the models.
They are two open-weight vision models that work as a pair. Image 001 creates and edits images, with deterministic seeds and style locks so results can be repeated. Detect 001 looks for diffusion fingerprints, inpainting seams and metadata mismatches, and returns a calibrated score with its confidence. Compare the two models.
Open-weight means the trained model parameters, the weights, are published so you can download and run the model yourself. It isn't the same as open-source, which also asks for the training code and a description of the training data. Our guide explains the difference.
Join the waitlist with your email address. You don't need an account to read this site. We email waitlist members when each model and its license are released.
No. A score is evidence, not proof. Use it alongside the image's source, provenance data such as content credentials, and human review, especially when the decision affects people.
No. The site sets no cookies and runs no analytics or advertising trackers.
Join the waitlist and we'll email you when the models and their licenses are released. One email, no spam, remove yourself any time.