A real person reads every message sent through this form. We aim to reply within five working days. If your message is a factual correction, it goes to the front of the queue — see our editorial policy for how corrections are handled.
What to contact us about
Report an error or request a correction
This is the message we most want to receive. Quote the sentence and give us the page URL, and we will verify it against the primary source. Confirmed factual errors are corrected in the page itself with an updated modification date — we do not quietly delete pages.
Technical questions about running GFPGAN
Installation failures, CUDA and PyTorch version conflicts, "model not found" errors in AUTOMATIC1111 or ComfyUI, or questions about which checkpoint suits your photos. Include your operating system, Python version, and the exact error text — it roughly triples the chance of a useful reply.
Privacy, data, and advertising
Access or deletion requests under GDPR or CCPA, questions about the cookies this site sets, or a report of intrusive advertising. We do not run popunders or interstitials, so if you saw one on this domain we want to know immediately.
Corrections about the model itself
If you work on GFPGAN or a related model and something here misrepresents your work, tell us. We will correct it promptly and credit the correction.
What we cannot help with
- We cannot restore photos for you as a service — the guides here show you how to do it yourself.
- We cannot provide support for commercial restoration apps we do not operate.
- We do not accept guest posts, link insertions, or paid placements in comparison articles. See our editorial policy.
- We are not Tencent ARC and cannot accept bug reports for GFPGAN itself — those belong on the project's GitHub issue tracker.
Email directly
If you would rather not use the form, email hello@gfpgaan.com. For privacy requests, please put "Privacy request" in the subject line so it is routed correctly.
Who you are contacting
GFPGAN.com is an independent documentation site covering GFPGAN and AI face restoration. We are not affiliated with Tencent ARC, who develop the GFPGAN model, and we do not represent them. Read more about us and about how we research and test.