←back to thread

Gemini 2.5 Flash Image

(developers.googleblog.com)
1092 points meetpateltech | 3 comments | 26 Aug 25 14:01 UTC | HN request time: 0.001s | source
Show context
vunderba ◴[26 Aug 25 16:49 UTC] No.45029113[source]▶
I've updated the GenAI Image comparison site (which focuses heavily on strict text-to-image prompt adherence) to reflect the new Google Gemini 2.5 Flash model (aka nano-banana).

https://genai-showdown.specr.net

This model gets 8 of the 12 prompts correct and easily comes within striking distance of the best-in-class models Imagen and gpt-image-1 and is a significant upgrade over the old Gemini Flash 2.0 model. The reigning champ, gpt-image-1, only manages to edge out Flash 2.5 on the maze and 9-pointed star.

What's honestly most astonishing to me is how long gpt-image-1 has remained at the top of the class - closing in on half a year which is basically a lifetime in this field. Though fair warning, gpt-image-1 is borderline useless as an "editor" since it almost always changes the whole image instead of doing localized inpainting-style edits like Kontext, Qwen, or Nano-Banana.

Comparison of gpt-image-1, flash, and imagen.

https://genai-showdown.specr.net?models=OPENAI_4O%2CIMAGEN_4...

replies(7): >>45030193 #>>45030194 #>>45030942 #>>45032937 #>>45033671 #>>45036899 #>>45041270 #
1. gundmc ◴[26 Aug 25 18:11 UTC] No.45030193[source]▶
> Though fair warning, gpt-image-1 is borderline useless as an "editor" since it almost always changes the whole image instead of doing localized inpainting-style edits like Kontext, Qwen, or Nano-Banana.

Came into this thread looking for this post. It's a great way to compare prompt adherence across models. Have you considered adding editing capabilities in a similar way given the recent trend of inpainting-style prompting?

replies(1): >>45031237 #
2. vunderba ◴[26 Aug 25 19:36 UTC] No.45031237[source]▶
Adding a separate section for image editing capabilities is a great idea.

I've done some experimentation with Qwen and Kontext and been pretty impressed, but it would be nice to see some side by sides now that we have essentially three models that are capable of highly localized in-painting without affecting the rest of the image.

https://mordenstar.com/blog/edits-with-kontext

replies(1): >>45050280 #
3. dostick ◴[28 Aug 25 09:49 UTC] No.45050280[source]▶
For editing prompts testing it is best to start with “only change …” to prevent model from changing everything. Even Nano banana does that.