I have searched the existing issues and discussions.
I am using the latest version of Open WebUI.
Installation Method
Pip Install
Open WebUI Version
0.6.13
Ollama Version (if applicable)
No response
Operating System
Ubuntu 24.04
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
I configured a MCP server using mcpo, which has about 8 tools inside it.
When user asked: is there available classrooms in T5 building tomorrow?
The LLM is supposed to:
get current time
get current school schedule (e.g. convert 2025.6.6 to 15th week)
use the schedule to fetch available classrooms
This involves 3 tools at least.
And I have tested this in Claude desktop as well as client like Cherry Studio. They all behaves well: generate a plan first, call the first tool, reflect on the result, and call the second tool... Finally it gives me the correct answer.
Actual Behavior
In Open WebUI, the LLM stops after calling the first tool, and then askes user that if they want to continue. Following that calling same tool for multiple times and then move to the next tool, with user clicking continue for multiple times.
I am sure that this is not related to the model itself as when I use the same model and the same prompt in Cherry Studio/Claude desktop, it never behaves like this.
Steps to Reproduce
The MCP server is hosted in school and only can be used when connected to school network. Sorry that I cannot give a reproduce step here.
Logs & Screenshots
I clicked continue multiple times here:
Additional Information
I wonder if there are user cases where multiple tools are called in a single round of conversation, and if LLM in Open WebUI can behave normally.
Originally created by @kowyo on GitHub (Jun 6, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/14714
### Check Existing Issues
- [x] I have searched the existing issues and discussions.
- [x] I am using the latest version of Open WebUI.
### Installation Method
Pip Install
### Open WebUI Version
0.6.13
### Ollama Version (if applicable)
_No response_
### Operating System
Ubuntu 24.04
### 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
I configured a MCP server using mcpo, which has about 8 tools inside it.
When user asked: is there available classrooms in T5 building tomorrow?
The LLM is supposed to:
- get current time
- get current school schedule (e.g. convert 2025.6.6 to 15th week)
- use the schedule to fetch available classrooms
This involves 3 tools at least.
And I have tested this in Claude desktop as well as client like Cherry Studio. They all behaves well: generate a plan first, call the first tool, reflect on the result, and call the second tool... Finally it gives me the correct answer.
### Actual Behavior
In Open WebUI, the LLM stops after calling the first tool, and then askes user that if they want to continue. Following that calling same tool for multiple times and then move to the next tool, with user clicking `continue` for multiple times.
I am sure that this is not related to the model itself as when I use the same model and the same prompt in Cherry Studio/Claude desktop, it never behaves like this.
### Steps to Reproduce
The MCP server is hosted in school and only can be used when connected to school network. Sorry that I cannot give a reproduce step here.
### Logs & Screenshots
I clicked continue multiple times here:
<img width="1019" alt="Image" src="https://github.com/user-attachments/assets/24825e88-d986-45c7-8fcf-fdb081a93dea" />
### Additional Information
I wonder if there are user cases where multiple tools are called in a single round of conversation, and if LLM in Open WebUI can behave normally.
GiteaMirror
added the bug label 2026-04-19 23:04:41 -05:00
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 @kowyo on GitHub (Jun 6, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/14714
Check Existing Issues
Installation Method
Pip Install
Open WebUI Version
0.6.13
Ollama Version (if applicable)
No response
Operating System
Ubuntu 24.04
Browser (if applicable)
No response
Confirmation
README.md.Expected Behavior
I configured a MCP server using mcpo, which has about 8 tools inside it.
When user asked: is there available classrooms in T5 building tomorrow?
The LLM is supposed to:
This involves 3 tools at least.
And I have tested this in Claude desktop as well as client like Cherry Studio. They all behaves well: generate a plan first, call the first tool, reflect on the result, and call the second tool... Finally it gives me the correct answer.
Actual Behavior
In Open WebUI, the LLM stops after calling the first tool, and then askes user that if they want to continue. Following that calling same tool for multiple times and then move to the next tool, with user clicking
continuefor multiple times.I am sure that this is not related to the model itself as when I use the same model and the same prompt in Cherry Studio/Claude desktop, it never behaves like this.
Steps to Reproduce
The MCP server is hosted in school and only can be used when connected to school network. Sorry that I cannot give a reproduce step here.
Logs & Screenshots
I clicked continue multiple times here:
Additional Information
I wonder if there are user cases where multiple tools are called in a single round of conversation, and if LLM in Open WebUI can behave normally.
@tjbck commented on GitHub (Jun 6, 2025):
Unable to reproduce, did you have the native mode toggled on?