github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 16:11:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #12879] Local Ollama Proxy sends Cloud model responses for non-streaming tasks with incorrect Content-Type header #55046

New Issue
Closed
opened 2026-04-29 08:14:46 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-04-29 08:14:46 -05:00
Owner
Copy Link

Originally created by @LosLockos13 on GitHub (Oct 31, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/12879

What is the issue?

Description:
When using a local Ollama instance as a proxy for Ollama Cloud models (models with the -cloud suffix like glm-4.6:cloud), the local instance returns the response from the cloud with a Content-Type: text/plain; charset=utf-8 header instead of the expected Content-Type: application/json.

This causes clients, like Open WebUI, to except with a ContentTypeError when trying to parse the valid JSON response.

Bypassing the local ollama instance and using https://ollama.com as backend does respond with correct Content-Type. So it looks like an issue in the local ollama instance when acting as proxy for cloud modles.

Steps to Reproduce:

  • Run a local Ollama instance.
  • Configure Open WebUI to use this local instance and select an cloud model like glm-4.6:cloud.
  • Trigger a background task in Open WebUI that uses a cloud model (e.g., automatic title generation, follow-up questions, or chat tagging)
  • Observe the error in the Open WebUI logs.

Expected Behavior:
The local Ollama instance, when proxying a request to the cloud, should preserve the Content-Type: application/json header from the cloud's response.

Actual Behavior:
The local Ollama instance delivers the JSON payload with a Content-Type: text/plain header.

Workaround:
A client-side workaround in Open WebUI is to ignore the Content-Type error for 200 OK responses and attempt to parse the body as JSON anyway.

Relevant log output

Oct 31 07:58:25 Host open-webui[47790]: > File "/srv/open-webui/lib/python3.11/site-packages/open_webui/routers/tasks.py", line 233, in generate_title
Oct 31 07:58:25 Host open-webui[47790]:     return await generate_chat_completion(request, form_data=payload, user=user)
Oct 31 07:58:25 Host open-webui[47790]:                  │                        │                  │             └ UserModel(id='3e9785c9-4a3f-4afb-89d2-103190d1eba2', name='xxx', email='xxx@home.net', username=None, role='admin', profile_i...
Oct 31 07:58:25 Host open-webui[47790]:                  │                        │                  └ {'model': 'glm-4.6:cloud', 'messages': [{'role': 'user', 'content': '### Task:\nGenerate a concise, 3-5 word title with an em...
Oct 31 07:58:25 Host open-webui[47790]:                  │                        └ <starlette.requests.Request object at 0x7fff31be4a50>
Oct 31 07:58:25 Host open-webui[47790]:                  └ <function generate_chat_completion at 0x7fff546ecf40>
Oct 31 07:58:25 Host open-webui[47790]:   File "/srv/open-webui/lib/python3.11/site-packages/open_webui/utils/chat.py", line 266, in generate_chat_completion
Oct 31 07:58:25 Host open-webui[47790]:     response = await generate_ollama_chat_completion(
Oct 31 07:58:25 Host open-webui[47790]:                      └ <function generate_chat_completion at 0x7fff547e2ac0>
Oct 31 07:58:25 Host open-webui[47790]:   File "/srv/open-webui/lib/python3.11/site-packages/open_webui/routers/ollama.py", line 1393, in generate_chat_completion
Oct 31 07:58:25 Host open-webui[47790]:     return await send_post_request(
Oct 31 07:58:25 Host open-webui[47790]:                  └ <function send_post_request at 0x7fff547653a0>
Oct 31 07:58:25 Host open-webui[47790]:   File "/srv/open-webui/lib/python3.11/site-packages/open_webui/routers/ollama.py", line 218, in send_post_request
Oct 31 07:58:25 Host open-webui[47790]:     raise HTTPException(
Oct 31 07:58:25 Host open-webui[47790]:           └ <class 'fastapi.exceptions.HTTPException'>
Oct 31 07:58:25 Host open-webui[47790]: fastapi.exceptions.HTTPException: 200: Ollama: 200, message='Attempt to decode JSON with unexpected mimetype: text/plain; charset=utf-8', url='http://Host:11434/api/chat'

OS

Linux

GPU

No response

CPU

Other

Ollama version

0.12.7

Originally created by @LosLockos13 on GitHub (Oct 31, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/12879 ### What is the issue? Description: When using a local Ollama instance as a proxy for Ollama Cloud models (models with the -cloud suffix like glm-4.6:cloud), the local instance returns the response from the cloud with a Content-Type: text/plain; charset=utf-8 header instead of the expected Content-Type: application/json. This causes clients, like Open WebUI, to except with a ContentTypeError when trying to parse the valid JSON response. Bypassing the local ollama instance and using https://ollama.com as backend does respond with correct Content-Type. So it looks like an issue in the local ollama instance when acting as proxy for cloud modles. Steps to Reproduce: - Run a local Ollama instance. - Configure Open WebUI to use this local instance and select an cloud model like glm-4.6:cloud. - Trigger a background task in Open WebUI that uses a cloud model (e.g., automatic title generation, follow-up questions, or chat tagging) - Observe the error in the Open WebUI logs. Expected Behavior: The local Ollama instance, when proxying a request to the cloud, should preserve the Content-Type: application/json header from the cloud's response. Actual Behavior: The local Ollama instance delivers the JSON payload with a Content-Type: text/plain header. Workaround: A client-side workaround in Open WebUI is to ignore the Content-Type error for 200 OK responses and attempt to parse the body as JSON anyway. ### Relevant log output ```shell Oct 31 07:58:25 Host open-webui[47790]: > File "/srv/open-webui/lib/python3.11/site-packages/open_webui/routers/tasks.py", line 233, in generate_title Oct 31 07:58:25 Host open-webui[47790]: return await generate_chat_completion(request, form_data=payload, user=user) Oct 31 07:58:25 Host open-webui[47790]: │ │ │ └ UserModel(id='3e9785c9-4a3f-4afb-89d2-103190d1eba2', name='xxx', email='xxx@home.net', username=None, role='admin', profile_i... Oct 31 07:58:25 Host open-webui[47790]: │ │ └ {'model': 'glm-4.6:cloud', 'messages': [{'role': 'user', 'content': '### Task:\nGenerate a concise, 3-5 word title with an em... Oct 31 07:58:25 Host open-webui[47790]: │ └ <starlette.requests.Request object at 0x7fff31be4a50> Oct 31 07:58:25 Host open-webui[47790]: └ <function generate_chat_completion at 0x7fff546ecf40> Oct 31 07:58:25 Host open-webui[47790]: File "/srv/open-webui/lib/python3.11/site-packages/open_webui/utils/chat.py", line 266, in generate_chat_completion Oct 31 07:58:25 Host open-webui[47790]: response = await generate_ollama_chat_completion( Oct 31 07:58:25 Host open-webui[47790]: └ <function generate_chat_completion at 0x7fff547e2ac0> Oct 31 07:58:25 Host open-webui[47790]: File "/srv/open-webui/lib/python3.11/site-packages/open_webui/routers/ollama.py", line 1393, in generate_chat_completion Oct 31 07:58:25 Host open-webui[47790]: return await send_post_request( Oct 31 07:58:25 Host open-webui[47790]: └ <function send_post_request at 0x7fff547653a0> Oct 31 07:58:25 Host open-webui[47790]: File "/srv/open-webui/lib/python3.11/site-packages/open_webui/routers/ollama.py", line 218, in send_post_request Oct 31 07:58:25 Host open-webui[47790]: raise HTTPException( Oct 31 07:58:25 Host open-webui[47790]: └ <class 'fastapi.exceptions.HTTPException'> Oct 31 07:58:25 Host open-webui[47790]: fastapi.exceptions.HTTPException: 200: Ollama: 200, message='Attempt to decode JSON with unexpected mimetype: text/plain; charset=utf-8', url='http://Host:11434/api/chat' ``` ### OS Linux ### GPU _No response_ ### CPU Other ### Ollama version 0.12.7
GiteaMirror added the bug label 2026-04-29 08:14:46 -05:00
GiteaMirror commented 2026-04-29 08:14:50 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Oct 31, 2025):

#12694

<!-- gh-comment-id:3472183691 --> @rick-github commented on GitHub (Oct 31, 2025): #12694
GiteaMirror commented 2026-04-29 08:14:51 -05:00
Author
Owner
Copy Link

@LosLockos13 commented on GitHub (Oct 31, 2025):

OK. I see. Thank you!

<!-- gh-comment-id:3472766873 --> @LosLockos13 commented on GitHub (Oct 31, 2025): OK. I see. Thank you!
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label bug
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#55046
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?