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
0.8.5
Ollama Version (if applicable)
No response
Operating System
ubuntu
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
With GPT-5-Mini selected and a custom tool that does search and scrape enabled in the workspae, in native mode it should work
Actual Behavior
the model returns that has not tools available.but if in Native mode it works
Steps to Reproduce
selecte gpt-5, register any custom tooll, enable and try to use in standard mode. I am using chat completions end point
Originally created by @MarceloMassarente on GitHub (Feb 24, 2026).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/21818
### 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
0.8.5
### Ollama Version (if applicable)
_No response_
### Operating System
ubuntu
### 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
With GPT-5-Mini selected and a custom tool that does search and scrape enabled in the workspae, in native mode it should work
### Actual Behavior
the model returns that has not tools available.but if in Native mode it works
### Steps to Reproduce
selecte gpt-5, register any custom tooll, enable and try to use in standard mode. I am using chat completions end point
### Logs & Screenshots
<img width="1348" height="720" alt="Image" src="https://github.com/user-attachments/assets/91ad6d9d-845d-4cb4-aa7c-135feb5d344e" />
26-02-24 05:42:20.953 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60809 - "GET /audio/notification.mp3 HTTP/1.1" 304
2026-02-24 05:42:21.120 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60809 - "GET /static/apple-touch-icon.png HTTP/1.1" 304
2026-02-24 05:42:21.401 | INFO | function_anthropic_api_new:get_anthropic_models:523 - Fetching models from Anthropic API...
2026-02-24 05:42:21.971 | INFO | function_anthropic_api_new:get_anthropic_models:549 - Successfully fetched 29 models from API (with fallback merge)
2026-02-24 05:42:22.344 | INFO | httpx._client:_send_single_request:1025 - HTTP Request: GET https://generativelanguage.googleapis.com/v1alpha/models "HTTP/1.1 200 OK"
2026-02-24 05:42:22.350 | INFO | open_webui.routers.openai:get_all_models:487 - get_all_models()
2026-02-24 05:42:22.357 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60810 - "GET /api/v1/chats/folder/736ef148-6f51-48fe-9128-1231cc3e4ce2/list?page=1 HTTP/1.1" 200
2026-02-24 05:42:22.403 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60811 - "GET /api/v1/tools/ HTTP/1.1" 200
2026-02-24 05:42:23.333 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60809 - "GET /api/models HTTP/1.1" 200
2026-02-24 05:42:40.029 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60814 - "GET /_app/version.json HTTP/1.1" 200
2026-02-24 05:43:00.348 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60825 - "POST /api/v1/chats/new HTTP/1.1" 200
2026-02-24 05:43:00.349 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60826 - "GET /api/v1/tools/ HTTP/1.1" 200
2026-02-24 05:43:00.417 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60825 - "GET /api/v1/chats/folder/736ef148-6f51-48fe-9128-1231cc3e4ce2/list?page=1 HTTP/1.1" 200
2026-02-24 05:43:00.446 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60827 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-02-24 05:43:00.505 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60826 - "POST /api/v1/chats/e2410d2e-e5a7-4c38-bc03-c37e7c2cf9a0 HTTP/1.1" 200
2026-02-24 05:43:00.533 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60826 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-02-24 05:43:00.588 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60826 - "POST /api/chat/completions HTTP/1.1" 200
2026-02-24 05:43:00.614 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60826 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-02-24 05:43:20.538 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:56327 - "POST /api/chat/completed HTTP/1.1" 200
2026-02-24 05:43:20.562 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:56327 - "POST /api/v1/chats/e2410d2e-e5a7-4c38-bc03-c37e7c2cf9a0 HTTP/1.1" 200
2026-02-24 05:43:20.583 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:56327 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-02-24 05:43:28.117 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:56328 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-02-24 05:43:28.137 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:56328 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-02-24 05:43:40.957 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:56329 - "GET /_app/version.json HTTP/1.1" 200
### Additional Information
_No response_
GiteaMirror
added the bug label 2026-05-05 22:38:54 -05:00
I see the same behavior since 0.8.4.
0.8.5 didn't fix it.
In my case the model doesn't see a tool that I use to create anki flashcards. Switching to native function calling works.
<!-- gh-comment-id:3950177614 -->
@timtips commented on GitHub (Feb 24, 2026):
I see the same behavior since 0.8.4.
0.8.5 didn't fix it.
In my case the model doesn't see a tool that I use to create anki flashcards. Switching to native function calling works.
I have the same problem on version 0.8.5. external tools with mcp are not working.
<!-- gh-comment-id:3950744056 -->
@arahnale commented on GitHub (Feb 24, 2026):
I have the same problem on version 0.8.5. external tools with mcp are not working.
Same, last working version for me is 0.8.3
Like @timtips since 0.8.4 my external tool are not called anymore
<!-- gh-comment-id:3950755436 -->
@mustard123 commented on GitHub (Feb 24, 2026):
Same, last working version for me is 0.8.3
Like @timtips since 0.8.4 my external tool are not called anymore
The same problem here, the internal and external tools worked fine in version 0.8.3. I upgraded to 0.8.5 and now there's no way for any model to call these tools.
<!-- gh-comment-id:3952295171 -->
@fauzcategui commented on GitHub (Feb 24, 2026):
The same problem here, the internal and external tools worked fine in version 0.8.3. I upgraded to 0.8.5 and now there's no way for any model to call these tools.
Same problem for me.
None of the models can call any tool now.
Not call failed, but they even do not know there is any tool available.
<!-- gh-comment-id:3952702056 -->
@lazyfai commented on GitHub (Feb 24, 2026):
Same problem for me.
None of the models can call any tool now.
Not call failed, but they even do not know there is any tool available.
can you reproduce it even on latest dev? can you guys please test latest dev? thanks
<!-- gh-comment-id:3953204896 -->
@Classic298 commented on GitHub (Feb 24, 2026):
can you reproduce it even on latest dev? can you guys please test latest dev? thanks
Same problem with version 0.8.5-slim.
dev-slim (git-b36e55c) and 0.8.3-slim is working.
<!-- gh-comment-id:3954600407 -->
@jmcomby commented on GitHub (Feb 24, 2026):
Same problem with version 0.8.5-slim.
dev-slim (git-b36e55c) and 0.8.3-slim is working.
Can you guys please confirm it's still present for you in dev or not? Thanks
<!-- gh-comment-id:3954644280 -->
@Classic298 commented on GitHub (Feb 24, 2026):
Can you guys please confirm it's still present for you in dev or not? Thanks
<!-- gh-comment-id:3954852118 -->
@fauzcategui commented on GitHub (Feb 24, 2026):
in latest dev it's working now
image: ghcr.io/open-webui/open-webui:dev
@AmanGupta0112 commented on GitHub (Feb 27, 2026):
Hi how did you fix the issue because i'm still not able to use the tools
I'm using this ghcr.io/open-webui/open-webui:dev but still it wont use any tool
<!-- gh-comment-id:3974525952 -->
@AmanGupta0112 commented on GitHub (Feb 27, 2026):
Hi how did you fix the issue because i'm still not able to use the tools
I'm using this ghcr.io/open-webui/open-webui:dev but still it wont use any tool
<img width="1728" height="1117" alt="Image" src="https://github.com/user-attachments/assets/cd07fe20-9bc6-432d-8af0-01c29a5e52b5" />
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 @MarceloMassarente on GitHub (Feb 24, 2026).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/21818
Check Existing Issues
Installation Method
Docker
Open WebUI Version
0.8.5
Ollama Version (if applicable)
No response
Operating System
ubuntu
Browser (if applicable)
No response
Confirmation
README.md.Expected Behavior
With GPT-5-Mini selected and a custom tool that does search and scrape enabled in the workspae, in native mode it should work
Actual Behavior
the model returns that has not tools available.but if in Native mode it works
Steps to Reproduce
selecte gpt-5, register any custom tooll, enable and try to use in standard mode. I am using chat completions end point
Logs & Screenshots
26-02-24 05:42:20.953 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60809 - "GET /audio/notification.mp3 HTTP/1.1" 304
2026-02-24 05:42:21.120 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60809 - "GET /static/apple-touch-icon.png HTTP/1.1" 304
2026-02-24 05:42:21.401 | INFO | function_anthropic_api_new:get_anthropic_models:523 - Fetching models from Anthropic API...
2026-02-24 05:42:21.971 | INFO | function_anthropic_api_new:get_anthropic_models:549 - Successfully fetched 29 models from API (with fallback merge)
2026-02-24 05:42:22.344 | INFO | httpx._client:_send_single_request:1025 - HTTP Request: GET https://generativelanguage.googleapis.com/v1alpha/models "HTTP/1.1 200 OK"
2026-02-24 05:42:22.350 | INFO | open_webui.routers.openai:get_all_models:487 - get_all_models()
2026-02-24 05:42:22.357 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60810 - "GET /api/v1/chats/folder/736ef148-6f51-48fe-9128-1231cc3e4ce2/list?page=1 HTTP/1.1" 200
2026-02-24 05:42:22.403 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60811 - "GET /api/v1/tools/ HTTP/1.1" 200
2026-02-24 05:42:23.333 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60809 - "GET /api/models HTTP/1.1" 200
2026-02-24 05:42:40.029 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60814 - "GET /_app/version.json HTTP/1.1" 200
2026-02-24 05:43:00.348 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60825 - "POST /api/v1/chats/new HTTP/1.1" 200
2026-02-24 05:43:00.349 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60826 - "GET /api/v1/tools/ HTTP/1.1" 200
2026-02-24 05:43:00.417 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60825 - "GET /api/v1/chats/folder/736ef148-6f51-48fe-9128-1231cc3e4ce2/list?page=1 HTTP/1.1" 200
2026-02-24 05:43:00.446 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60827 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-02-24 05:43:00.505 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60826 - "POST /api/v1/chats/e2410d2e-e5a7-4c38-bc03-c37e7c2cf9a0 HTTP/1.1" 200
2026-02-24 05:43:00.533 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60826 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-02-24 05:43:00.588 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60826 - "POST /api/chat/completions HTTP/1.1" 200
2026-02-24 05:43:00.614 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:60826 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-02-24 05:43:20.538 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:56327 - "POST /api/chat/completed HTTP/1.1" 200
2026-02-24 05:43:20.562 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:56327 - "POST /api/v1/chats/e2410d2e-e5a7-4c38-bc03-c37e7c2cf9a0 HTTP/1.1" 200
2026-02-24 05:43:20.583 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:56327 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-02-24 05:43:28.117 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:56328 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-02-24 05:43:28.137 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:56328 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-02-24 05:43:40.957 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:56329 - "GET /_app/version.json HTTP/1.1" 200
Additional Information
No response
@timtips commented on GitHub (Feb 24, 2026):
I see the same behavior since 0.8.4.
0.8.5 didn't fix it.
In my case the model doesn't see a tool that I use to create anki flashcards. Switching to native function calling works.
@arahnale commented on GitHub (Feb 24, 2026):
I have the same problem on version 0.8.5. external tools with mcp are not working.
@mustard123 commented on GitHub (Feb 24, 2026):
Same, last working version for me is 0.8.3
Like @timtips since 0.8.4 my external tool are not called anymore
@fauzcategui commented on GitHub (Feb 24, 2026):
The same problem here, the internal and external tools worked fine in version 0.8.3. I upgraded to 0.8.5 and now there's no way for any model to call these tools.
@lazyfai commented on GitHub (Feb 24, 2026):
Same problem for me.
None of the models can call any tool now.
Not call failed, but they even do not know there is any tool available.
@marktemi commented on GitHub (Feb 24, 2026):
yes same issue
@Classic298 commented on GitHub (Feb 24, 2026):
can you reproduce it even on latest dev? can you guys please test latest dev? thanks
@jmcomby commented on GitHub (Feb 24, 2026):
Same problem with version 0.8.5-slim.
dev-slim (git-b36e55c) and 0.8.3-slim is working.
@Classic298 commented on GitHub (Feb 24, 2026):
Can you guys please confirm it's still present for you in dev or not? Thanks
@fauzcategui commented on GitHub (Feb 24, 2026):
in latest dev it's working now
image: ghcr.io/open-webui/open-webui:dev
@Classic298 commented on GitHub (Feb 24, 2026):
thanks for confirming. And with that: fixed in dev.
@lazyfai commented on GitHub (Feb 25, 2026):
Dev docker image seems resolved the issue for new chats (only).
@AmanGupta0112 commented on GitHub (Feb 27, 2026):
Hi how did you fix the issue because i'm still not able to use the tools
I'm using this ghcr.io/open-webui/open-webui:dev but still it wont use any tool