issue: Cannot generate more than 1 image in a chat

[tolidano]

Check Existing Issues

• I have searched for any existing and/or related issues.
• I have searched for any existing and/or related discussions.
• I have also searched in the CLOSED issues AND CLOSED discussions and found no related items (your issue might already be addressed on the development branch!).
• I am using the latest version of Open WebUI.

Installation Method

Pip Install

Open WebUI Version

0.7.2

Ollama Version (if applicable)

LM Studio

Operating System

macOS Tahoe

Browser (if applicable)

Safari

Confirmation

• I have read and followed all instructions in README.md.
• I am using the latest version of both Open WebUI and Ollama.
• I have included the browser console logs.
• I have included the Docker container logs.
• I have provided every relevant configuration, setting, and environment variable used in my setup.
• I have clearly listed every relevant configuration, custom setting, environment variable, and command-line option that influences my setup (such as Docker Compose overrides, .env values, browser settings, authentication configurations, etc).
• I have documented step-by-step reproduction instructions that are precise, sequential, and leave nothing to interpretation. My steps:
• Start with the initial platform/version/OS and dependencies used,
• Specify exact install/launch/configure commands,
• List URLs visited, user input (incl. example values/emails/passwords if needed),
• Describe all options and toggles enabled or changed,
• Include any files or environmental changes,
• Identify the expected and actual result at each stage,
• Ensure any reasonably skilled user can follow and hit the same issue.

Expected Behavior

When I ask it to generate an image, it works the first time in a chat, then on subsequent requests, it should also work.

Actual Behavior

We need to say error.I’m sorry, but I couldn’t generate the image you requested. The following error occurred: [ERROR: Expecting value: line 1 column 1 (char 0)]

After the first image, it fails on every subsequent image with the same error. If I delete this chat and start a new one, the first image works again.

Nothing in the logs in ComfyUI or in Open WebUI or in LM Studio logs that indicate the reason for the error.

Steps to Reproduce

Clean install open web UI, lm studio (with gpt-oss-20b), comfyUI (with z-image turbo template workflow).

Start a chat
Ask for an image
See image successfully
Ask for another image
See error
Start new chat
Ask for an image
See image successfully

Logs & Screenshots

2026-01-18 12:38:11.888 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:49993 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-01-18 12:38:11.916 | INFO     | open_webui.routers.openai:get_all_models:477 - get_all_models()
2026-01-18 12:38:11.917 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:49992 - "POST /api/chat/completions HTTP/1.1" 200
2026-01-18 12:38:11.928 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:49993 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-01-18 12:38:14.149 | INFO     | open_webui.utils.images.comfyui:comfyui_create_image:198 - WebSocket connection established.
2026-01-18 12:38:14.149 | INFO     | open_webui.utils.images.comfyui:comfyui_create_image:204 - Sending workflow to WebSocket server.
2026-01-18 12:38:14.149 | INFO     | open_webui.utils.images.comfyui:comfyui_create_image:205 - Workflow: {'9': {'inputs': {'filename_prefix': 'z-image', 'images': ['43', 0]}, 'class_type': 'SaveImage', '_meta': {'
title': 'Save Image'}}, '39': {'inputs': {'clip_name': 'qwen_3_4b.safetensors', 'type': 'lumina2', 'device': 'default'}, 'class_type': 'CLIPLoader', '_meta': {'title': 'Load CLIP'}}, '40': {'inputs': {'vae_name': '
ae.safetensors'}, 'class_type': 'VAELoader', '_meta': {'title': 'Load VAE'}}, '41': {'inputs': {'width': 512, 'height': 512, 'batch_size': 1}, 'class_type': 'EmptySD3LatentImage', '_meta': {'title': 'EmptySD3Latent
Image'}}, '42': {'inputs': {'conditioning': ['45', 0]}, 'class_type': 'ConditioningZeroOut', '_meta': {'title': 'ConditioningZeroOut'}}, '43': {'inputs': {'samples': ['44', 0], 'vae': ['40', 0]}, 'class_type': 'VAE
Decode', '_meta': {'title': 'VAE Decode'}}, '44': {'inputs': {'seed': 237843675715867, 'steps': 30, 'cfg': 1, 'sampler_name': 'res_multistep', 'scheduler': 'simple', 'denoise': 1, 'model': ['47', 0], 'positive': ['
45', 0], 'negative': ['42', 0], 'latent_image': ['41', 0]}, 'class_type': 'KSampler', '_meta': {'title': 'KSampler'}}, '45': {'inputs': {'text': "A playful, cartoon-style illustration featuring a friendly dinosaur
with a large, rounded body and a small tail. The dinosaur has a cheerful expression, bright green scales with slightly darker shading on the sides to suggest depth. Its front legs are short and stubby, holding a fr
eshly baked blueberry muffin that sits on its belly. The muffin is golden-brown with a slightly textured crumb and specks of dark blueberries scattered across its top. The dinosaur’s mouth is open in a wide, deligh
ted grin as it takes a big bite, crumbs falling around its cheeks. The background is a simple pastel blue sky with a few fluffy white clouds, giving the scene a light, whimsical atmosphere. The overall composition
is bright and kid-friendly, with bold outlines and smooth color transitions typical of children's cartoons. ", 'clip': ['39', 0]}, 'class_type': 'CLIPTextEncode', '_meta': {'title': 'CLIP Text Encode (Prompt)'}}, '
46': {'inputs': {'unet_name': 'z_image_turbo_bf16.safetensors', 'weight_dtype': 'default'}, 'class_type': 'UNETLoader', '_meta': {'title': 'Load Diffusion Model'}}, '47': {'inputs': {'shift': 3, 'model': ['46', 0]}
, 'class_type': 'ModelSamplingAuraFlow', '_meta': {'title': 'ModelSamplingAuraFlow'}}}
2026-01-18 12:38:14.150 | INFO     | open_webui.utils.images.comfyui:queue_prompt:20 - queue_prompt
2026-01-18 12:38:56.827 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:50071 - "GET /_app/version.json HTTP/1.1" 304
2026-01-18 12:39:03.546 | INFO     | open_webui.utils.images.comfyui:get_history:57 - get_history
2026-01-18 12:39:03.547 | INFO     | open_webui.utils.images.comfyui:get_image_url:50 - get_image
2026-01-18 12:39:03.549 | INFO     | open_webui.routers.files:upload_file_handler:225 - file.content_type: image/png False
2026-01-18 12:39:03.555 | INFO     | open_webui.routers.openai:get_all_models:477 - get_all_models()
2026-01-18 12:39:03.578 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:50080 - "GET /api/v1/files/055aa743-77cf-499e-90be-645c102892af/content HTTP/1.1" 200
2026-01-18 12:39:04.595 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:50080 - "POST /api/chat/completed HTTP/1.1" 200
2026-01-18 12:39:04.618 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:50080 - "POST /api/v1/chats/a208ede3-e8d1-4dc7-b7e8-688f89124ac8 HTTP/1.1" 200
2026-01-18 12:39:04.716 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:50080 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200

Additional Information

No response

[owui-terminator[bot]]

🔍 Similar Issues Found

I found some existing issues that might be related to this one. Please check if any of these are duplicates or contain helpful solutions:

1. #20436 issue: image generation cannot finish normally
   by lywbh • Jan 07, 2026 • bug

2. #18995 issue: image generation and edition doesn’t work on temporary chats
   by futureshield • Nov 06, 2025 • bug

3. #16953 issue: When switching to another chat while the image is being generated, the original chat cannot display the generated image properly.
   by haochiu • Aug 27, 2025 • bug

4. #19393 issue: shared chats with images - images won't show
   by Classic298 • Nov 23, 2025 • bug, confirmed issue

5. #20091 issue: image is regarded as binary in temp chat
   by funnycups • Dec 22, 2025 • bug

---

💡 Tips:

• If this is a duplicate, please consider closing this issue and adding any additional details to the existing one
• If you found a solution in any of these issues, please share it here to help others

This comment was generated automatically by a bot. Please react with a 👍 if this comment was helpful, or a 👎 if it was not.

[Classic298]

We're gonna need a hell of a lot more details here, i.e. full setup details for image generation, env vars and more.

As for me, i cannot reproduce this with a normal setup nor with image-native models.

[tolidano]

No env vars, I literally just do open-webui serve after pipx install open-webui (with the latest 0.7.2, I did inject graphviz as well, fwiw). Here are my ComfyUI settings in open webui

here are my ComfyUI startup (I just cloned it and ran it, following the standard setup instructions for PyTorch nightly to run on my Studio M2 Ultra 192GB:

❯ p main.py
Checkpoint files will always be loaded safely.
Total VRAM 196608 MB, total RAM 196608 MB
pytorch version: 2.9.1
Mac Version (26, 2)
Set vram state to: SHARED
Device: mps
Using sub quadratic optimization for attention, if you have memory or speed issues try using: --use-split-cross-attention
Python version: 3.13.11 (main, Dec 24 2025, 12:12:07) [Clang 17.0.0 (clang-1700.6.3.2)]
ComfyUI version: 0.6.0
ComfyUI frontend version: 1.35.9
[Prompt Server] web root: /Users/shawn/.pyenv/versions/3.13.11/lib/python3.13/site-packages/comfyui_frontend_package/static
Total VRAM 196608 MB, total RAM 196608 MB
pytorch version: 2.9.1
Mac Version (26, 2)
Set vram state to: SHARED
Device: mps

Import times for custom nodes:
   0.0 seconds: /Users/shawn/dev/ComfyUI/custom_nodes/websocket_image_save.py

Context impl SQLiteImpl.
Will assume non-transactional DDL.
No target revision found.
Starting server

To see the GUI go to: http://127.0.0.1:8188
got prompt
Using split attention in VAE
Using split attention in VAE
VAE load device: mps, offload device: cpu, dtype: torch.bfloat16
Requested to load ZImageTEModel_
loaded completely; 95367431640625005117571072.00 MB usable, 7672.25 MB loaded, full load: True
CLIP/text encoder model load device: cpu, offload device: cpu, current: cpu, dtype: torch.float16
model weight dtype torch.bfloat16, manual cast: None
model_type FLOW

As noted, the first image works perfectly.

Here's a reddit thread with two other folks having the same problem, if that's relevant: https://www.reddit.com/r/OpenWebUI/comments/1oy60jr/image_generation_issues/

[Classic298]

Ah ComfyUI. @silentoplayz can you help here?

[silentoplayz]

Sure! First off, Z-Image-Turbo is a turbo model, "a distilled version of Z-Image that matches or exceeds leading competitors with only 8 NFEs (Number of Function Evaluations)". OP should drop the steps count from 30 to 8, tops.

The second mistake I see OP making here is a missing node ID for his image generation workflow connected to Open WebUI, specifically for ckpt_name. Each node ID must be filled out accordingly to the uploaded image generation workflow.

e.g.:

It also looks like OP attempted to edit an image using a text prompt after an image was generated, but failed to provide a screenshot of the image edit settings, so I could only help regarding image generation here.

[tolidano]

Thank you for these suggestions. Can you share why these would result in needing a brand new chat to generate a completely novel image?

I will make the corrections noted and post my other failing test, to generate two completely different images in the same chat.

[tolidano]

The workflow (from the z-image turbo template) has no field for chkpt_name in the steps, and it is not in the JSON (which would be expected, if it isn't in the steps).

I did change the steps to 8 as you suggested.

Here I am with a brand new chat:

And then right after:

Logs on Open Web UI:

2026-01-19 12:26:21.902 | INFO     | open_webui.utils.images.comfyui:comfyui_create_image:204 - Sending workflow to WebSocket server.
2026-01-19 12:26:21.903 | INFO     | open_webui.utils.images.comfyui:comfyui_create_image:205 - Workflow: {'9': {'inputs': {'filename_prefix': 'z-image', 'images': ['43', 0]}, 'class_type': 'SaveImage', '_meta': {'
title': 'Save Image'}}, '39': {'inputs': {'clip_name': 'qwen_3_4b.safetensors', 'type': 'lumina2', 'device': 'default'}, 'class_type': 'CLIPLoader', '_meta': {'title': 'Load CLIP'}}, '40': {'inputs': {'vae_name': '
ae.safetensors'}, 'class_type': 'VAELoader', '_meta': {'title': 'Load VAE'}}, '41': {'inputs': {'width': 1024, 'height': 1024, 'batch_size': 1}, 'class_type': 'EmptySD3LatentImage', '_meta': {'title': 'EmptySD3Late
ntImage'}}, '42': {'inputs': {'conditioning': ['45', 0]}, 'class_type': 'ConditioningZeroOut', '_meta': {'title': 'ConditioningZeroOut'}}, '43': {'inputs': {'samples': ['44', 0], 'vae': ['40', 0]}, 'class_type': 'V
AEDecode', '_meta': {'title': 'VAE Decode'}}, '44': {'inputs': {'seed': 714196730802686, 'steps': 8, 'cfg': 1, 'sampler_name': 'res_multistep', 'scheduler': 'simple', 'denoise': 1, 'model': ['47', 0], 'positive': [
'45', 0], 'negative': ['42', 0], 'latent_image': ['41', 0]}, 'class_type': 'KSampler', '_meta': {'title': 'KSampler'}}, '45': {'inputs': {'text': 'draw a picture of a bunny please', 'clip': ['39', 0]}, 'class_type'
: 'CLIPTextEncode', '_meta': {'title': 'CLIP Text Encode (Prompt)'}}, '46': {'inputs': {'unet_name': 'z_image_turbo_bf16.safetensors', 'weight_dtype': 'default'}, 'class_type': 'UNETLoader', '_meta': {'title': 'Loa
d Diffusion Model'}}, '47': {'inputs': {'shift': 3, 'model': ['46', 0]}, 'class_type': 'ModelSamplingAuraFlow', '_meta': {'title': 'ModelSamplingAuraFlow'}}}
2026-01-19 12:26:21.903 | INFO     | open_webui.utils.images.comfyui:queue_prompt:20 - queue_prompt
2026-01-19 12:26:37.898 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61366 - "GET /audio/notification.mp3 HTTP/1.1" 206
2026-01-19 12:26:37.906 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61366 - "GET /audio/notification.mp3 HTTP/1.1" 206
2026-01-19 12:26:37.932 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61366 - "GET /static/apple-touch-icon.png HTTP/1.1" 304
2026-01-19 12:26:37.994 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61366 - "GET /static/favicon-96x96.png HTTP/1.1" 200
2026-01-19 12:26:38.001 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61366 - "GET /static/favicon.svg HTTP/1.1" 200
2026-01-19 12:26:38.013 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61366 - "GET /static/favicon.ico HTTP/1.1" 200
2026-01-19 12:26:39.325 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61366 - "GET /_app/version.json HTTP/1.1" 304
2026-01-19 12:27:13.606 | INFO     | open_webui.utils.images.comfyui:get_history:57 - get_history
2026-01-19 12:27:13.607 | INFO     | open_webui.utils.images.comfyui:get_image_url:50 - get_image
2026-01-19 12:27:13.609 | INFO     | open_webui.routers.files:upload_file_handler:225 - file.content_type: image/png False
2026-01-19 12:27:13.616 | INFO     | open_webui.routers.openai:get_all_models:477 - get_all_models()
2026-01-19 12:27:13.639 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61426 - "GET /api/v1/files/e12b2c38-db96-47ee-9857-429e1ad30efb/content HTTP/1.1" 200
2026-01-19 12:27:20.378 | INFO     | open_webui.routers.openai:get_all_models:477 - get_all_models()
2026-01-19 12:27:20.414 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61445 - "POST /api/chat/completed HTTP/1.1" 200
2026-01-19 12:27:20.432 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61445 - "POST /api/v1/chats/67706f5d-c12a-4caf-ba82-23f7ee657976 HTTP/1.1" 200
2026-01-19 12:27:20.445 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61445 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-01-19 12:27:21.713 | INFO     | open_webui.routers.openai:get_all_models:477 - get_all_models()
2026-01-19 12:27:22.724 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61445 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-01-19 12:27:23.468 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61445 - "GET /api/v1/chats/67706f5d-c12a-4caf-ba82-23f7ee657976 HTTP/1.1" 200
2026-01-19 12:27:23.481 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61445 - "GET /api/v1/chats/all/tags HTTP/1.1" 200
2026-01-19 12:27:27.616 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61445 - "POST /api/v1/chats/67706f5d-c12a-4caf-ba82-23f7ee657976 HTTP/1.1" 200
2026-01-19 12:27:27.628 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61445 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-01-19 12:27:27.655 | INFO     | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.198:61445 - "POST /api/chat/completions HTTP/1.1" 200
2026-01-19 12:27:27.663 | INFO     | open_webui.routers.openai:get_all_models:477 - get_all_models()

Logs in ComfyUI:

got prompt
100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 8/8 [00:50<00:00,  6.32s/it]
Prompt executed in 51.56 seconds

Note: just for the bunny, no logs received for the parrot request

Logs in LM Studio (where I host oss-20b):

2026-01-19 12:27:27 [DEBUG]
 Received request: POST to /v1/chat/completions with body  {
  "stream": true,
  "model": "openai/gpt-oss-20b",
  "messages": [
    {
      "role": "system",
      "content": "<context>Image generation was attempted but failed... <Truncated in logs> ...pecting value: line 1 column 1 (char 0)]</context>"
    },
    {
      "role": "user",
      "content": "draw a picture of a bunny please"
    },
    {
      "role": "assistant",
      "content": "\nYour requested picture of a bunny has been created and is now displayed in the chat. Enjoy!"
    },
    {
      "role": "user",
      "content": "now draw a picture of a parrot please."
    }
  ]
}
2026-01-19 12:27:27  [INFO]
 [openai/gpt-oss-20b] Running chat completion on conversation with 4 messages.
2026-01-19 12:27:27  [INFO]
 [openai/gpt-oss-20b] Streaming response...
2026-01-19 12:27:27 [DEBUG]
 Prompt successfully formatted with Harmony.
2026-01-19 12:27:27 [DEBUG]
 Prompt successfully formatted with Harmony.
2026-01-19 12:27:27 [DEBUG]
 [cache_wrapper][WARNING]: Tried to trim '227' tokens from the prompt cache, but could not: Cache is not trimmable. Clearing the cache instead.
2026-01-19 12:27:27  [INFO]
 [openai/gpt-oss-20b] Prompt processing progress: 0.0%
2026-01-19 12:27:27  [INFO]
 [openai/gpt-oss-20b] Prompt processing progress: 100.0%
2026-01-19 12:27:28  [INFO]
 [openai/gpt-oss-20b] Finished streaming response
2026-01-19 12:27:28 [DEBUG]
 Received request: POST to /v1/chat/completions with body  {
  "model": "openai/gpt-oss-20b",
  "messages": [
    {
      "role": "user",
      "content": "### Task:\nSuggest 3-5 relevant follow-up questions... <Truncated in logs> ...g value: line 1 column 1 (char 0)]\n</chat_history>"
    }
  ],
  "stream": false
}

Note system context now has the failure for the image (prior to this, it's just the generation of the bunny, the titling, followups, categories, etc.

ED: Also worth noting I immediately clicked "New chat" and asked for the same parrot and it appeared promptly.

[tolidano]

Any other possible recommendations? Any other config or logs to share?