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
Docker
Open WebUI Version
0.9.2
Ollama Version (if applicable)
N/A
Operating System
Ubuntu 24.04
Browser (if applicable)
Chrome
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 chatting with LLM, I expect the analytics dashboard to show estimated token usage for my current and only user who is also the admin.
Actual Behavior
Analytics dashboard shows 0 tokens but does count the # of messages correctly when connected to AWS bedrock via gateway. Seems to work okay with Ollama on my local machine.
Steps to Reproduce
Running OpenWebUI via Docker for years
Connect with AWS Bedrock via self-hosted gateway
Recently migrated to OpenWebUI from 0.92 from 0.7.2.
No other issues so far.
Select an Amazon model (tried Kimi K2.5, Anthropic's Sonnet 2.6)
Start chatting with LLM
Load analytics dashboard and observe 0 count for tokens. Message count works.
Observe that chat works perfectly otherwise and I can interact freely with AWS-hosted model.
Logs & Screenshots
Additional Information
I was surprised that OpenWebUI didn't have more robust information on token usage. Even the text generation and prompt processing statistics are a bit hard to read and don't always match counts I've observed in llama.cpp (console).
For this issue, since I'm using a paid API via AWS, it would be helpful if the UI could provide me with real time accounting of usage that's reasonably accurate (+/- 1K tokens). I didn't see bug report for the analytics dashboard but there is a thread on this feature that's open.
Happy to provide logs from self-hosted bedrock gateway. Thanks.
Originally created by @d-shehu on GitHub (Apr 29, 2026).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/24241
### Check Existing Issues
- [x] I have searched for any existing and/or related issues.
- [x] I have searched for any existing and/or related discussions.
- [x] 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!).
- [x] I am using the latest version of Open WebUI.
### Installation Method
Docker
### Open WebUI Version
0.9.2
### Ollama Version (if applicable)
N/A
### Operating System
Ubuntu 24.04
### Browser (if applicable)
Chrome
### 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 chatting with LLM, I expect the analytics dashboard to show estimated token usage for my current and only user who is also the admin.
### Actual Behavior
Analytics dashboard shows 0 tokens but does count the # of messages correctly when connected to AWS bedrock via gateway. Seems to work okay with Ollama on my local machine.
### Steps to Reproduce
1. Running OpenWebUI via Docker for years
2. Connect with AWS Bedrock via self-hosted gateway
3. Recently migrated to OpenWebUI from 0.92 from 0.7.2.
* No other issues so far.
4. Select an Amazon model (tried Kimi K2.5, Anthropic's Sonnet 2.6)
5. Start chatting with LLM
6. Load analytics dashboard and observe 0 count for tokens. Message count works.
7. Observe that chat works perfectly otherwise and I can interact freely with AWS-hosted model.
### Logs & Screenshots
<img width="1636" height="536" alt="Image" src="https://github.com/user-attachments/assets/1de0eb31-68c7-4802-919f-d25b162f304a" />
### Additional Information
I was surprised that OpenWebUI didn't have more robust information on token usage. Even the text generation and prompt processing statistics are a bit hard to read and don't always match counts I've observed in llama.cpp (console).
For this issue, since I'm using a paid API via AWS, it would be helpful if the UI could provide me with real time accounting of usage that's reasonably accurate (+/- 1K tokens). I didn't see bug report for the analytics dashboard but there is a thread on this feature that's open.
Happy to provide logs from self-hosted bedrock gateway. Thanks.
GiteaMirror
added the bug label 2026-05-15 16:16:28 -05:00
<!-- gh-comment-id:4374930568 -->
@Classic298 commented on GitHub (May 4, 2026):
likely fixed by https://github.com/open-webui/open-webui/commit/989d5fd4e2ce285edf4475a1e13f0981a78d3821 and https://github.com/open-webui/open-webui/commit/a32d26e61d24d9f63650faed5cb8909ed90af661
Testing wanted on dev
Thanks, I still would tag this so it gets retested when the release has those fixes. Don't see the reason for prematurely closing tickets.
And I'm happy to retest but I've got enough to tackle without dealing with pre-release builds. Sorry.
<!-- gh-comment-id:4375214859 -->
@d-shehu commented on GitHub (May 4, 2026):
Thanks, I still would tag this so it gets retested when the release has those fixes. Don't see the reason for prematurely closing tickets.
And I'm happy to retest but I've got enough to tackle without dealing with pre-release builds. Sorry.
yes it will be tagged in the changelogs. we close it because we use tickets for... well tracking. so this is closed because should be fixed.
users are told to search opened and closed issues either way before opening one
<!-- gh-comment-id:4375221812 -->
@Classic298 commented on GitHub (May 4, 2026):
yes it will be tagged in the changelogs. we close it because we use tickets for... well tracking. so this is closed because should be fixed.
users are told to search opened and closed issues either way before opening one
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 @d-shehu on GitHub (Apr 29, 2026).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/24241
Check Existing Issues
Installation Method
Docker
Open WebUI Version
0.9.2
Ollama Version (if applicable)
N/A
Operating System
Ubuntu 24.04
Browser (if applicable)
Chrome
Confirmation
README.md.Expected Behavior
When chatting with LLM, I expect the analytics dashboard to show estimated token usage for my current and only user who is also the admin.
Actual Behavior
Analytics dashboard shows 0 tokens but does count the # of messages correctly when connected to AWS bedrock via gateway. Seems to work okay with Ollama on my local machine.
Steps to Reproduce
Running OpenWebUI via Docker for years
Connect with AWS Bedrock via self-hosted gateway
Recently migrated to OpenWebUI from 0.92 from 0.7.2.
Select an Amazon model (tried Kimi K2.5, Anthropic's Sonnet 2.6)
Start chatting with LLM
Load analytics dashboard and observe 0 count for tokens. Message count works.
Observe that chat works perfectly otherwise and I can interact freely with AWS-hosted model.
Logs & Screenshots
Additional Information
I was surprised that OpenWebUI didn't have more robust information on token usage. Even the text generation and prompt processing statistics are a bit hard to read and don't always match counts I've observed in llama.cpp (console).
For this issue, since I'm using a paid API via AWS, it would be helpful if the UI could provide me with real time accounting of usage that's reasonably accurate (+/- 1K tokens). I didn't see bug report for the analytics dashboard but there is a thread on this feature that's open.
Happy to provide logs from self-hosted bedrock gateway. Thanks.
@masterturk commented on GitHub (May 1, 2026):
I have the same issue.
@Classic298 commented on GitHub (May 4, 2026):
likely fixed by https://github.com/open-webui/open-webui/commit/989d5fd4e2ce285edf4475a1e13f0981a78d3821 and https://github.com/open-webui/open-webui/commit/a32d26e61d24d9f63650faed5cb8909ed90af661
Testing wanted on dev
@d-shehu commented on GitHub (May 4, 2026):
Thanks, I still would tag this so it gets retested when the release has those fixes. Don't see the reason for prematurely closing tickets.
And I'm happy to retest but I've got enough to tackle without dealing with pre-release builds. Sorry.
@Classic298 commented on GitHub (May 4, 2026):
yes it will be tagged in the changelogs. we close it because we use tickets for... well tracking. so this is closed because should be fixed.
users are told to search opened and closed issues either way before opening one