hoffmanngabriel89Q: Selfie Upload Limits, Full‑Body AI Capture & Model Combination
I’m interested in your product and have a few questions:
1. Number of Selfie Uploads
How many photos can I upload at most to train my “clone”? Some YouTube videos mention only a few uploads. Since more high‑quality images improve results, is there an upper limit and do you plan to raise it?
2. Full‑Body Capture
Can the AI learn not only facial features but also hands, fingers, and the rest of the body in detail? I want not only a realistic portrait but also a coherent full‑body representation with anatomically correct fingers.
3. Combining Multiple Models
Is it possible to combine two pre‑trained models in one image? For example, myself and another person interacting in the same scene.
Please let me know within two days so I can still take advantage of the AppSumo discount.
Thank you and kind regards,
Gabriel Hoffmann
Creative_Bot_PhotoGPTAI
May 20, 2025A: Hi Gabriel, thanks so much for your thoughtful questions, happy to help clarify!
- Number of Selfie Uploads
Currently, the limit for model training is 12–15 photos. This range has proven to offer the best balance between consistency and flexibility, enabling the model to capture facial features accurately while allowing diverse image generations across themes. That said, based on user demand, we’re working on increasing this limit to 20 images in the coming days — stay tuned!
- Full-Body Capture
The PhotoGPT AI platform is optimized to deliver highly accurate facial reproduction, ideal for photorealistic headshots. You can, however, train the model using only full-body images if you’d like it to generate outputs showing full poses. Just note that in such cases, the facial fidelity may reduce slightly since the training will be more broadly focused across the body.
- Combining Multiple Models
By default, our models are optimized for a single subject. However, we do have a beta feature that allows training a two-character model for scenes involving interactions. You can learn more about this method and how to set it up via our guide here:
👉 https://www.photogptai.com/guides/two-character-model-training-(beta)
Let us know if you have more questions — we’d love to have you onboard before the AppSumo deal ends!
Thank you!
Three questions:
1. From which deal are all future features and updates included? Does this apply starting at Tier 2 or only at a higher plan?
2. Does PhotoGPT adapt to technological advancements in image AI models? And what new features are currently planned?
3. What AI technology is the image model based on? Is it a variation of Stable Diffusion, Flux, or a proprietary Model?