I have searched the existing issues and discussions.
I am using the latest version of Open WebUI.
Installation Method
Git Clone
Open WebUI Version
v0.6.16
Ollama Version (if applicable)
No response
Operating System
Windows 10
Browser (if applicable)
No response
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
Standard users are able to correctly use the knowledge created by admins.
Actual Behavior
Standard users are able to include knowledge/collections in their chats, but any request involving the knowledge/collections included are answered saying something like that no context was provided (we tested several models, and they all have the same behaviour).
Please note that if an admin user sends exactly the same message with the same knowledge/collections, it gets an answer relevant to the context provided.
Steps to Reproduce
Login as an admin user
Load some knowledge in a collection
Open a new chat
Include that collection
Ask a question in a new chat regarding that knowledge
you should get a reasonable answer
Login as a standard user
Open a new chat
Include that same collection
Ask exactly the same question asked in 5
the model should say something like: no context provided
Logs & Screenshots
Unavailable
Additional Information
I needed to check all the checkboxes above to be able to open this issue, but the following are false in this case:
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.
Originally created by @daniporr on GitHub (Jul 17, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/15801
### Check Existing Issues
- [x] I have searched the existing issues and discussions.
- [x] I am using the latest version of Open WebUI.
### Installation Method
Git Clone
### Open WebUI Version
v0.6.16
### Ollama Version (if applicable)
_No response_
### Operating System
Windows 10
### Browser (if applicable)
_No response_
### 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
Standard users are able to correctly use the knowledge created by admins.
### Actual Behavior
Standard users are able to include knowledge/collections in their chats, but any request involving the knowledge/collections included are answered saying something like that no context was provided (we tested several models, and they all have the same behaviour).
Please note that if an admin user sends exactly the same message with the same knowledge/collections, it gets an answer relevant to the context provided.
### Steps to Reproduce
1. Login as an admin user
2. Load some knowledge in a collection
3. Open a new chat
4. Include that collection
5. Ask a question in a new chat regarding that knowledge
- you should get a reasonable answer
6. Login as a standard user
7. Open a new chat
8. Include that same collection
9. Ask exactly the same question asked in 5
- the model should say something like: _no context provided_
### Logs & Screenshots
Unavailable
### Additional Information
I needed to check all the checkboxes above to be able to open this issue, but the following are false in this case:
- 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.
GiteaMirror
added the bug label 2026-04-25 07:06:52 -05:00
I can not replicate that in v0.16. I created a knowledge base as an admin, which has 3 documents, one is a nvidia financial report (many pages and numbers, good for testing). The the answer to the question "What was the financial gain of nvidia last year?" included the source in both cases, admin and normal user.
<!-- gh-comment-id:3084330814 -->
@mahenning commented on GitHub (Jul 17, 2025):
I can not replicate that in v0.16. I created a knowledge base as an admin, which has 3 documents, one is a nvidia financial report (many pages and numbers, good for testing). The the answer to the question "What was the financial gain of nvidia last year?" included the source in both cases, admin and normal user.
I upgraded to the latest version and now my knowledge base is broken. Seems to be an issue with reranking according to logs.
I assume knowledge base != knowledge base in this case.
Rolling back to previous release instantly fixes the issue for me.
Will create a separate issue later.
<!-- gh-comment-id:3085411054 -->
@apunkt commented on GitHub (Jul 17, 2025):
I upgraded to the latest version and now my knowledge base is broken. Seems to be an issue with reranking according to logs.
I assume knowledge base != knowledge base in this case.
Rolling back to previous release instantly fixes the issue for me.
Will create a separate issue later.
When standard user click a document in citation (in a model with knowledge created by admin) and try to open it, get this error:
{"detail":"We could not find what you're looking for :/"}
<!-- gh-comment-id:3090593703 -->
@galvanoid commented on GitHub (Jul 18, 2025):
Similar issue here.
When standard user click a document in citation (in a model with knowledge created by admin) and try to open it, get this error:
{"detail":"We could not find what you're looking for :/"}
@spammenotinoz commented on GitHub (Jul 27, 2025):
Yes indeed, however in my case the issue occured when;
User is using a MODEL created by an ADMIN and KNOWLEDGE they created.
Using the files individually within the collection works, however referencing the Collection, the LLM has no knowledge.
Admins do not have the issue.
<!-- gh-comment-id:3124725486 -->
@spammenotinoz commented on GitHub (Jul 27, 2025):
Yes indeed, however in my case the issue occured when;
- User is using a MODEL created by an ADMIN and KNOWLEDGE they created.
Using the files individually within the collection works, however referencing the Collection, the LLM has no knowledge.
Admins do not have the issue.
Yes indeed, however in my case the issue occured when;
User is using a MODEL created by an ADMIN and KNOWLEDGE they created.
Using the files individually within the collection works, however referencing the Collection, the LLM has no knowledge. Admins do not have the issue.
I can reproduce this scenario. I created a model with admin and logged in with a non-admin user. Added a collection with that user then used this collection with admin created model. Model couldn't find the context.
<!-- gh-comment-id:3125890483 -->
@ctcanbol commented on GitHub (Jul 28, 2025):
> Yes indeed, however in my case the issue occured when;
>
> * User is using a MODEL created by an ADMIN and KNOWLEDGE they created.
>
> Using the files individually within the collection works, however referencing the Collection, the LLM has no knowledge. Admins do not have the issue.
I can reproduce this scenario. I created a model with admin and logged in with a non-admin user. Added a collection with that user then used this collection with admin created model. Model couldn't find the context.
Regular users are unable to use notes in normal chat, they can select the notes but they are not added to the chat for LLM, other user created documents in the knowledge base seem to work fine.
<!-- gh-comment-id:3127423084 -->
@IgorFZ commented on GitHub (Jul 28, 2025):
Version 0.6.18:
Regular users are unable to use notes in normal chat, they can select the notes but they are not added to the chat for LLM, other user created documents in the knowledge base seem to work fine.
It works normally for admin users.
https://github.com/user-attachments/assets/8f31fd3a-04ba-4e92-a16a-5d2e4b710ed5
Yes indeed, however in my case the issue occured when;
User is using a MODEL created by an ADMIN and KNOWLEDGE they created.
Using the files individually within the collection works, however referencing the Collection, the LLM has no knowledge. Admins do not have the issue.
I can reproduce this scenario. I created a model with admin and logged in with a non-admin user. Added a collection with that user then used this collection with admin created model. Model couldn't find the context.
@tjbck is this also fixed in dev?
<!-- gh-comment-id:3159816942 -->
@ctcanbol commented on GitHub (Aug 6, 2025):
> > Yes indeed, however in my case the issue occured when;
> >
> > * User is using a MODEL created by an ADMIN and KNOWLEDGE they created.
> >
> > Using the files individually within the collection works, however referencing the Collection, the LLM has no knowledge. Admins do not have the issue.
>
> I can reproduce this scenario. I created a model with admin and logged in with a non-admin user. Added a collection with that user then used this collection with admin created model. Model couldn't find the context.
@tjbck is this also fixed in dev?
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 @daniporr on GitHub (Jul 17, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/15801
Check Existing Issues
Installation Method
Git Clone
Open WebUI Version
v0.6.16
Ollama Version (if applicable)
No response
Operating System
Windows 10
Browser (if applicable)
No response
Confirmation
README.md.Expected Behavior
Standard users are able to correctly use the knowledge created by admins.
Actual Behavior
Standard users are able to include knowledge/collections in their chats, but any request involving the knowledge/collections included are answered saying something like that no context was provided (we tested several models, and they all have the same behaviour).
Please note that if an admin user sends exactly the same message with the same knowledge/collections, it gets an answer relevant to the context provided.
Steps to Reproduce
Logs & Screenshots
Unavailable
Additional Information
I needed to check all the checkboxes above to be able to open this issue, but the following are false in this case:
@mahenning commented on GitHub (Jul 17, 2025):
I can not replicate that in v0.16. I created a knowledge base as an admin, which has 3 documents, one is a nvidia financial report (many pages and numbers, good for testing). The the answer to the question "What was the financial gain of nvidia last year?" included the source in both cases, admin and normal user.
@apunkt commented on GitHub (Jul 17, 2025):
I upgraded to the latest version and now my knowledge base is broken. Seems to be an issue with reranking according to logs.
I assume knowledge base != knowledge base in this case.
Rolling back to previous release instantly fixes the issue for me.
Will create a separate issue later.
@galvanoid commented on GitHub (Jul 18, 2025):
Similar issue here.
When standard user click a document in citation (in a model with knowledge created by admin) and try to open it, get this error:
{"detail":"We could not find what you're looking for :/"}
@spammenotinoz commented on GitHub (Jul 27, 2025):
Yes indeed, however in my case the issue occured when;
Using the files individually within the collection works, however referencing the Collection, the LLM has no knowledge.
Admins do not have the issue.
@ctcanbol commented on GitHub (Jul 28, 2025):
I can reproduce this scenario. I created a model with admin and logged in with a non-admin user. Added a collection with that user then used this collection with admin created model. Model couldn't find the context.
@IgorFZ commented on GitHub (Jul 28, 2025):
Version 0.6.18:
Regular users are unable to use notes in normal chat, they can select the notes but they are not added to the chat for LLM, other user created documents in the knowledge base seem to work fine.
It works normally for admin users.
https://github.com/user-attachments/assets/8f31fd3a-04ba-4e92-a16a-5d2e4b710ed5
@tjbck commented on GitHub (Aug 4, 2025):
@IgorFZ addressed note attachment issue in dev.
@spammenotinoz commented on GitHub (Aug 6, 2025):
Thank-you
@ctcanbol commented on GitHub (Aug 6, 2025):
@tjbck is this also fixed in dev?