-
Notifications
You must be signed in to change notification settings - Fork 680
Open
Description
I'm using genkit js and the plugin @genkit-ai/google-genai
I tried generating a 0.5K image, but it is totally impossible. It always returns a 1K image.
The documentation is not very clear about the parameter we have to send to create a 512px image:
https://ai.google.dev/gemini-api/docs/image-generation
Gemini 3 image models generate 1K images by default but can also output 2K, 4K, and 512px (05.K) (Gemini 3.1 Flash Image only) images. To generate higher resolution assets, specify the image_size in the generation_config.
You must use an uppercase 'K' (e.g. 512px (05.K), 1K, 2K, 4K). Lowercase parameters (e.g., 1k) will be rejected.
I tried "512px", "0.5K", and "05.K", but nothing, it always returns a 1024 image.
I used curl and all those possible values. None of the worked. So, I assume the problem is in the model itself. (or google apis)
Do you see the same issue?
curl -s -X POST \
"https://generativelanguage.googleapis.com/v1beta/models/gemini-3.1-flash-image-preview:generateContent" \
-H "x-goog-api-key: XXXXXXXX" \
-H "Content-Type: application/json" \
-d '{
"contents": [{
"parts": [
{"text": "Create a picture of a nano banana dish in a fancy restaurant with a Gemini theme"}
]
}],
"generationConfig": {
"responseModalities": ["IMAGE"],
"imageConfig": {
"imageSize": "0.5K",
"aspectRatio": "1:1"
}
}
}' | python3 -c "
import sys, json, base64
data = json.load(sys.stdin)
for part in data['candidates'][0]['content']['parts']:
if 'inlineData' in part:
img = base64.b64decode(part['inlineData']['data'])
with open('/tmp/test.png', 'wb') as f:
f.write(img)
print(f'Saved to /tmp/test.png ({len(img)} bytes)')
break
"
EDIT: I just saw this:
https://discuss.ai.google.dev/t/imagesize-512px-not-applied-in-gemini-3-1-flash-image-preview-always-returns-1k-resolution/127332
Reactions are currently unavailable
Metadata
Metadata
Assignees
Labels
No labels
Type
Projects
Status
No status