I have searched the existing issues and discussions.
I am using the latest version of Open WebUI.
Installation Method
Docker
Open WebUI Version
v0.6.16
Ollama Version (if applicable)
N/A
Operating System
Windows 11
Browser (if applicable)
Edge
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 files were uploaded to the server, their content should not be resubmitted with each new chat message. Instead, it should be extracted internally from storage.
Actual Behavior
Each new chat message resends the entire content of each uploaded/attached file. Since some files are ~500KB in size - chat message size snowballs out of control quickly and "nginx" balancer starts to throw 413 errors.
Steps to Reproduce
Enable "Paste Large Text as File" in interface settings
Enable "Bypass Embedding and Retrieval" in Document admin settings
"CTRL + V" a large text content in new chat - observe that it is successfully uploaded as file
Submit the message. Observe that the POST request contains the entire file content!
Originally created by @R0ll1ngSt0ne on GitHub (Jul 18, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/15846
### Check Existing Issues
- [x] I have searched the existing issues and discussions.
- [x] I am using the latest version of Open WebUI.
### Installation Method
Docker
### Open WebUI Version
v0.6.16
### Ollama Version (if applicable)
N/A
### Operating System
Windows 11
### Browser (if applicable)
Edge
### Confirmation
- [x] I have read and followed all instructions in `README.md`.
- [x] I am using the latest version of **both** Open WebUI and Ollama.
- [x] I have included the browser console logs.
- [x] I have included the Docker container logs.
- [x] I have **provided every relevant configuration, setting, and environment variable used in my setup.**
- [x] 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).
- [x] 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 files were uploaded to the server, their content should not be resubmitted with each new chat message. Instead, it should be extracted internally from storage.
### Actual Behavior
Each new chat message resends the entire content of each uploaded/attached file. Since some files are ~500KB in size - chat message size snowballs out of control quickly and "nginx" balancer starts to throw 413 errors.
### Steps to Reproduce
1. Enable "Paste Large Text as File" in interface settings
2. Enable "Bypass Embedding and Retrieval" in Document admin settings
3. "CTRL + V" a large text content in new chat - observe that it is successfully uploaded as file
4. Submit the message. Observe that the POST request contains the entire file content!
### Logs & Screenshots
[llmt-experimentation-dev.ai-nprod.intactfc.cloud.txt](https://github.com/user-attachments/files/21321486/llmt-experimentation-dev.ai-nprod.intactfc.cloud.txt)
<img width="887" height="243" alt="Image" src="https://github.com/user-attachments/assets/9ad1debb-c1b2-4cff-a6b0-fdf7253dde34" />
<img width="749" height="268" alt="Image" src="https://github.com/user-attachments/assets/d4e50a97-5802-4c13-bdcd-a459b6d72175" />
### Additional Information
_No response_
GiteaMirror
added the bug label 2026-04-25 07:07:23 -05:00
But why? It is a huge network bandwidth waste to keep re-uploading attached files for every submitted message - why not just reuse the uploaded content?
<!-- gh-comment-id:3096895983 -->
@R0ll1ngSt0ne commented on GitHub (Jul 21, 2025):
> Intended behaviour.
But why? It is a huge network bandwidth waste to keep re-uploading attached files for every submitted message - why not just reuse the uploaded content?
because that is a task that have to be done in the model provider endpoint, and right now that function is not in its APIs,
e.g. a RETAIN_CONTEXT_IN_CONVERSATION=true param & a conversationID param, then in model provider side it know that it have to keep in memory messsages of the conversation and that the stream only have new message....
but for this it have to be in the API provider, right now, no openAI, no ollama, have it.
<!-- gh-comment-id:3096989392 -->
@rgaricano commented on GitHub (Jul 21, 2025):
because that is a task that have to be done in the model provider endpoint, and right now that function is not in its APIs,
e.g. a RETAIN_CONTEXT_IN_CONVERSATION=true param & a conversationID param, then in model provider side it know that it have to keep in memory messsages of the conversation and that the stream only have new message....
but for this it have to be in the API provider, right now, no openAI, no ollama, have it.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Originally created by @R0ll1ngSt0ne on GitHub (Jul 18, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/15846
Check Existing Issues
Installation Method
Docker
Open WebUI Version
v0.6.16
Ollama Version (if applicable)
N/A
Operating System
Windows 11
Browser (if applicable)
Edge
Confirmation
README.md.Expected Behavior
When files were uploaded to the server, their content should not be resubmitted with each new chat message. Instead, it should be extracted internally from storage.
Actual Behavior
Each new chat message resends the entire content of each uploaded/attached file. Since some files are ~500KB in size - chat message size snowballs out of control quickly and "nginx" balancer starts to throw 413 errors.
Steps to Reproduce
Logs & Screenshots
llmt-experimentation-dev.ai-nprod.intactfc.cloud.txt
Additional Information
No response
@tjbck commented on GitHub (Jul 19, 2025):
Intended behaviour.
@R0ll1ngSt0ne commented on GitHub (Jul 21, 2025):
But why? It is a huge network bandwidth waste to keep re-uploading attached files for every submitted message - why not just reuse the uploaded content?
@rgaricano commented on GitHub (Jul 21, 2025):
because that is a task that have to be done in the model provider endpoint, and right now that function is not in its APIs,
e.g. a RETAIN_CONTEXT_IN_CONVERSATION=true param & a conversationID param, then in model provider side it know that it have to keep in memory messsages of the conversation and that the stream only have new message....
but for this it have to be in the API provider, right now, no openAI, no ollama, have it.