Image generation only produces 1 output + editing fails
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 39.5k
- Forks
- 4.8k
- PR merge metrics
- No merged PRs in 30d
Description
Hello, I’m experiencing repeated issues with image generation on LMArena in the past days:
1. When I upload a photo, the model only generates one output image, even if the prompt or model previously produced multiple variations.
2. After generating the first image, editing no longer works. Any attempt to edit or re-generate triggers an “Error” message.
3. Some images load only partially and go into infinite buffering, or the generation fails immediately.
4. This issue didn’t exist before — the same models (especially Nano-Banana / Flash) used to create multiple variations with no errors.
5. I tested different browsers, devices, new chat sessions, and clearing cache. The issue persists, so it seems server-side.
Models affected:
- Nano-Banana
- Flash image editing models
- Possibly others
Steps to reproduce:
1. Start a new chat
2. Upload a photo
3. Generate
4. Attempt to generate more than one image or edit the output
5. Error appears consistently
Expected behavior:
Multiple image variations and successful editing (as previously available).
Actual behavior:
Only one image generated; further edits or generations fail.
Thanks for checking this — the issue appears to be server-side and consistent for many users.
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
No source file or test is identified in the report. Start by reproducing the upload, generation, and editing steps in a new chat, then trace the server-side path for Nano-Banana and Flash image models. Done means multiple image variations can be generated, edits succeed, and images no longer fail or buffer indefinitely.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- ai, backend
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100