Low-resolution camera stills from trail cams, doorbell cameras, and bird feeders have always been a pain to work with. Dedicated upscaling tools like Topaz Photo AI exist, but they cost $199 and up. Turns out ChatGPT's image generation can handle basic photo enhancement for free.
Users have been feeding grainy, low-quality photos into ChatGPT and asking it to "improve the quality," and the results are genuinely usable. The AI doesn't just sharpen pixels the way traditional upscalers do. Because it's working with a generative model (meaning it creates new image data rather than just interpolating between existing pixels), it can fill in plausible detail that wasn't in the original. Feathers get texture. Backgrounds gain depth. The overall image looks like it came from a much better camera.
The catch: this is generation, not restoration. ChatGPT is making educated guesses about what those blurry pixels should look like. Fine details like text on a sign or the exact pattern on a bird's wing might not match reality. For scientific documentation or legal evidence, stick with proper forensic tools. For sharing a cool bird photo or making a security camera still look presentable, it works surprisingly well.
This is a free capability available to all ChatGPT users through the standard image generation feature. No plugins, no special prompts. Just upload a photo and ask for an improved version. The results won't replace Photoshop for professional retouching, but for quick cleanup of low-quality images, it's hard to argue with free and instant.