[GH-ISSUE #5 ] issue: OpenTerminal not working #5

Owner

Originally created by @MarceloMassarente on GitHub (Feb 23, 2026).
Original GitHub issue: https://github.com/open-webui/open-terminal/issues/5

Check Existing Issues

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

select gpt-5, native mode, call openterminal and get a response

Actual Behavior

Invalid schema for function 'read_file': None is not valid under any of the given schemas.

Steps to Reproduce

just try Openterminal in native mode

Logs & Screenshots

026-02-23 20:09:11.630 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:63698 - "GET /api/models HTTP/1.1" 200
2026-02-23 20:09:18.790 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:63709 - "GET /_app/version.json HTTP/1.1" 200
2026-02-23 20:09:55.231 | ERROR | open_webui.socket.main:periodic_session_pool_cleanup:187 - Unable to renew session cleanup lock. Exiting.
2026-02-23 20:09:56.207 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52071 - "POST /api/tasks/stop/07134e6b-c04a-4f01-bfa2-bf7ccebb8b46 HTTP/1.1" 200
2026-02-23 20:09:58.683 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52071 - "POST /api/v1/chats/f776bcac-cc26-46a2-baea-2a7ea85263a7 HTTP/1.1" 200
2026-02-23 20:09:58.710 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52071 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-02-23 20:09:58.764 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52071 - "POST /api/chat/completions HTTP/1.1" 200
2026-02-23 20:09:58.791 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52071 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-02-23 20:10:19.768 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52073 - "GET /_app/version.json HTTP/1.1" 200
2026-02-23 20:10:46.120 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52075 - "POST /api/tasks/stop/4116130f-cac6-473c-8b04-3eb772996f58 HTTP/1.1" 200
2026-02-23 20:10:56.780 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52078 - "POST /api/v1/chats/f776bcac-cc26-46a2-baea-2a7ea85263a7 HTTP/1.1" 200
2026-02-23 20:10:56.806 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52078 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-02-23 20:10:56.864 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52078 - "POST /api/chat/completions HTTP/1.1" 200
2026-02-23 20:10:56.894 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52078 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200
2026-02-23 20:11:20.839 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52001 - "GET /_app/version.json HTTP/1.1" 200
2026-02-23 20:12:21.774 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:65273 - "GET /_app/version.json HTTP/1.1" 200

Additional Information

No response

Originally created by @MarceloMassarente on GitHub (Feb 23, 2026). Original GitHub issue: https://github.com/open-webui/open-terminal/issues/5 ### 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 select gpt-5, native mode, call openterminal and get a response ### Actual Behavior Invalid schema for function 'read_file': None is not valid under any of the given schemas. ### Steps to Reproduce just try Openterminal in native mode ### Logs & Screenshots 026-02-23 20:09:11.630 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:63698 - "GET /api/models HTTP/1.1" 200 2026-02-23 20:09:18.790 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:63709 - "GET /_app/version.json HTTP/1.1" 200 2026-02-23 20:09:55.231 | ERROR | open_webui.socket.main:periodic_session_pool_cleanup:187 - Unable to renew session cleanup lock. Exiting. 2026-02-23 20:09:56.207 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52071 - "POST /api/tasks/stop/07134e6b-c04a-4f01-bfa2-bf7ccebb8b46 HTTP/1.1" 200 2026-02-23 20:09:58.683 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52071 - "POST /api/v1/chats/f776bcac-cc26-46a2-baea-2a7ea85263a7 HTTP/1.1" 200 2026-02-23 20:09:58.710 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52071 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200 2026-02-23 20:09:58.764 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52071 - "POST /api/chat/completions HTTP/1.1" 200 2026-02-23 20:09:58.791 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52071 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200 2026-02-23 20:10:19.768 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52073 - "GET /_app/version.json HTTP/1.1" 200 2026-02-23 20:10:46.120 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52075 - "POST /api/tasks/stop/4116130f-cac6-473c-8b04-3eb772996f58 HTTP/1.1" 200 2026-02-23 20:10:56.780 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52078 - "POST /api/v1/chats/f776bcac-cc26-46a2-baea-2a7ea85263a7 HTTP/1.1" 200 2026-02-23 20:10:56.806 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52078 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200 2026-02-23 20:10:56.864 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52078 - "POST /api/chat/completions HTTP/1.1" 200 2026-02-23 20:10:56.894 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52078 - "GET /api/v1/chats/?page=1 HTTP/1.1" 200 2026-02-23 20:11:20.839 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:52001 - "GET /_app/version.json HTTP/1.1" 200 2026-02-23 20:12:21.774 | INFO | uvicorn.protocols.http.httptools_impl:send:483 - 192.168.1.230:65273 - "GET /_app/version.json HTTP/1.1" 200 ### Additional Information _No response_

bsakaguchi

closed this issue
added the
bug
label

bsakaguchi referenced this issue

[GH-ISSUE #4] Bug Report: gpt-5 returns error of "invalid schema for function 'read_file': None is not valid under any of the given schemas." #4

bsakaguchi commented

Author

Owner

@MarceloMassarente commented on GitHub (Feb 23, 2026):

Invalid schema for function 'read_file': None is not valid under any of the given schemas.

in Standard mode it is not even recognized

@MarceloMassarente commented on GitHub (Feb 23, 2026): <img width="1652" height="391" alt="Image" src="https://github.com/user-attachments/assets/fb71854b-3297-4ffa-8f81-f9148e55a013" /> Invalid schema for function 'read_file': None is not valid under any of the given schemas. in Standard mode it is not even recognized

bsakaguchi commented

Author

Owner

@tjbck commented on GitHub (Feb 24, 2026):

Is this using azure ai?

@tjbck commented on GitHub (Feb 24, 2026): Is this using azure ai?

bsakaguchi commented

Author

Owner

@tjbck commented on GitHub (Feb 24, 2026):

@silenceroom @MarceloMassarente could anyone confirm the issue has been resolved in the latest dev (open-webui)?

@tjbck commented on GitHub (Feb 24, 2026): @silenceroom @MarceloMassarente could anyone confirm the issue has been resolved in the latest dev (open-webui)?

bsakaguchi commented

Author

Owner

@MarceloMassarente commented on GitHub (Feb 24, 2026):

hi, this happens with regular open ai key

@MarceloMassarente commented on GitHub (Feb 24, 2026): hi, this happens with regular open ai key

bsakaguchi commented

Author

Owner

@Seviks commented on GitHub (Feb 24, 2026):

i can confirm this happens on v.0.8.5 with gpt-5.2 (regular openai) with native mode.

same error with dev (open-webui) pulled just now.

@Seviks commented on GitHub (Feb 24, 2026): i can confirm this happens on v.0.8.5 with gpt-5.2 (regular openai) with native mode. same error with dev (open-webui) pulled just now.

bsakaguchi commented

Author

Owner

@tjbck commented on GitHub (Feb 24, 2026):

@Seviks I'm unable to reproduce the issue with the latest open-webui and open-terminal, could you share your screen recording with us?

@tjbck commented on GitHub (Feb 24, 2026): @Seviks I'm unable to reproduce the issue with the latest open-webui and open-terminal, could you share your screen recording with us?

bsakaguchi commented

Author

Owner

@cloudtuotuo commented on GitHub (Feb 25, 2026):

We're using New-API as the backend. I just pulled newest open-terminal docker few minitues ago. Our open-webui is on v0.8.5.
When using gpt models, claude models and deepseek models through New-API, the same issue occured. When using ollama models, no issues.

The logs from New-API:
new-api | [ERR] 2026/02/25 - 09:24:44 | 20260225012444239828895ck0CRJLq | channel error (channel #1, status code: 400): Invalid schema for function 'read_file': None is not valid under any of the given schemas.
new-api | [INFO] 2026/02/25 - 09:24:44 | 20260225012444239828895ck0CRJLq | record error log: userId=1, channelId=1, modelName=gpt-5.2, tokenName=OpenWebUI, content=status_code=400, Invalid schema for function 'read_file': None is not valid under any of the given schemas.
new-api | [ERR] 2026/02/25 - 09:24:44 | 20260225012444239828895ck0CRJLq | relay error: Invalid schema for function 'read_file': None is not valid under any of the given schemas.

new-api | [ERR] 2026/02/25 - 09:31:59 | 20260225013158965884000K5U3yisq | channel error (channel #5, status code: 400): at '/properties/start_line/type': value must be one of 'array', 'boolean', 'integer', 'null', 'number', 'object', 'string'
new-api | [INFO] 2026/02/25 - 09:31:59 | 20260225013158965884000K5U3yisq | record error log: userId=1, channelId=5, modelName=deepseek-v3, tokenName=OpenWebUI, content=status_code=400, at '/properties/start_line/type': value must be one of 'array', 'boolean', 'integer', 'null', 'number', 'object', 'string'
new-api | [ERR] 2026/02/25 - 09:31:59 | 20260225013158965884000K5U3yisq | relay error: at '/properties/start_line/type': value must be one of 'array', 'boolean', 'integer', 'null', 'number', 'object', 'string'

@cloudtuotuo commented on GitHub (Feb 25, 2026): We're using New-API as the backend. I just pulled newest open-terminal docker few minitues ago. Our open-webui is on v0.8.5. When using gpt models, claude models and deepseek models through New-API, the same issue occured. When using ollama models, no issues. The logs from New-API: new-api | [ERR] 2026/02/25 - 09:24:44 | 20260225012444239828895ck0CRJLq | channel error (channel #1, status code: 400): Invalid schema for function 'read_file': None is not valid under any of the given schemas. new-api | [INFO] 2026/02/25 - 09:24:44 | 20260225012444239828895ck0CRJLq | record error log: userId=1, channelId=1, modelName=gpt-5.2, tokenName=OpenWebUI, content=status_code=400, Invalid schema for function 'read_file': None is not valid under any of the given schemas. new-api | [ERR] 2026/02/25 - 09:24:44 | 20260225012444239828895ck0CRJLq | relay error: Invalid schema for function 'read_file': None is not valid under any of the given schemas. new-api | [ERR] 2026/02/25 - 09:31:59 | 20260225013158965884000K5U3yisq | channel error (channel #5, status code: 400): at '/properties/start_line/type': value must be one of 'array', 'boolean', 'integer', 'null', 'number', 'object', 'string' new-api | [INFO] 2026/02/25 - 09:31:59 | 20260225013158965884000K5U3yisq | record error log: userId=1, channelId=5, modelName=deepseek-v3, tokenName=OpenWebUI, content=status_code=400, at '/properties/start_line/type': value must be one of 'array', 'boolean', 'integer', 'null', 'number', 'object', 'string' new-api | [ERR] 2026/02/25 - 09:31:59 | 20260225013158965884000K5U3yisq | relay error: at '/properties/start_line/type': value must be one of 'array', 'boolean', 'integer', 'null', 'number', 'object', 'string'

bsakaguchi commented

https://github.com/user-attachments/assets/1e33056c-6a25-4867-a21c-35abb4976a43

Author

Owner

@silenceroom commented on GitHub (Feb 25, 2026):

Is this using azure ai?

yes, it's Azure AI on my case. I can confirm the issue remains with the v0.8.5 and the latest open_terminal.

@silenceroom commented on GitHub (Feb 25, 2026): > Is this using azure ai? yes, it's Azure AI on my case. I can confirm the issue remains with the v0.8.5 and the latest open_terminal. https://github.com/user-attachments/assets/1e33056c-6a25-4867-a21c-35abb4976a43

bsakaguchi commented

Author

Owner

@Seviks commented on GitHub (Feb 25, 2026):

With the latest changes the dev container (all clean) does not start (Error when running migrations, but i deleted volumes). I'll try again tomorrow.

@silenceroom I think we're meant to test with latest owui and latest open-terminal.

@Seviks commented on GitHub (Feb 25, 2026): With the latest changes the dev container (all clean) does not start (Error when running migrations, but i deleted volumes). I'll try again tomorrow. @silenceroom I think we're meant to test with latest owui **and** latest open-terminal.

bsakaguchi commented

Author

Owner

@silenceroom commented on GitHub (Feb 25, 2026):

With the latest changes the dev container (all clean) does not start (Error when running migrations, but i deleted volumes). I'll try again tomorrow.

@silenceroom I think we're meant to test with latest owui and latest open-terminal.

same problem here, the latest dev image does not even start with the container.

@silenceroom commented on GitHub (Feb 25, 2026): > With the latest changes the dev container (all clean) does not start (Error when running migrations, but i deleted volumes). I'll try again tomorrow. > > [@silenceroom](https://github.com/silenceroom) I think we're meant to test with latest owui **and** latest open-terminal. same problem here, the latest dev image does not even start with the container.

bsakaguchi commented

Author

Owner

@tjbck commented on GitHub (Feb 25, 2026):

@silenceroom @Seviks @MarceloMassarente we're unable to reproduce with the lastest dev with gpt-5.2 (native function calling enabled)

@tjbck commented on GitHub (Feb 25, 2026): @silenceroom @Seviks @MarceloMassarente we're unable to reproduce with the lastest dev with gpt-5.2 (native function calling enabled)

bsakaguchi commented

Author

Owner

@silenceroom commented on GitHub (Feb 26, 2026):

@silenceroom @Seviks @MarceloMassarente we're unable to reproduce with the lastest dev with gpt-5.2 (native function calling enabled)

I can confirm the problem remains in the latest dev docker image. I'll run the source code later to see if it's fixed.

@silenceroom commented on GitHub (Feb 26, 2026): > [@silenceroom](https://github.com/silenceroom) [@Seviks](https://github.com/Seviks) [@MarceloMassarente](https://github.com/MarceloMassarente) we're unable to reproduce with the lastest dev with gpt-5.2 (native function calling enabled) I can confirm the problem remains in the latest dev docker image. I'll run the source code later to see if it's fixed.

bsakaguchi commented

Author

Owner

@Seviks commented on GitHub (Feb 26, 2026):

Maybe I/we do something wrong in the config. I can reproduce the error with this example:

Run default containers in in a docker compose:

services:

  open-webui:
    container_name: open-webui
    image: ghcr.io/open-webui/open-webui:dev
    ports:
      - 3000:8080
    restart: unless-stopped
    volumes:
      - open-webui:/app/backend/data


  open-terminal:
    image: ghcr.io/open-webui/open-terminal
    container_name: open-terminal
    volumes:
      - open-terminal:/home/user
    ports:
      - 8000:8000
    environment:
      - OPEN_TERMINAL_API_KEY=sk-aaa

volumes:
  open-terminal:
  open-webui:

Verify that owui reaches open-terminal: docker exec -it open-webui curl http://open-terminal:8000/docs
Create login, add OpenAI-API-Key
Configure Tool:

5. Add Tool to Chat with native mode:

@Seviks commented on GitHub (Feb 26, 2026): Maybe I/we do something wrong in the config. I can reproduce the error with this example: 1. Run default containers in in a docker compose: ``` services: open-webui: container_name: open-webui image: ghcr.io/open-webui/open-webui:dev ports: - 3000:8080 restart: unless-stopped volumes: - open-webui:/app/backend/data open-terminal: image: ghcr.io/open-webui/open-terminal container_name: open-terminal volumes: - open-terminal:/home/user ports: - 8000:8000 environment: - OPEN_TERMINAL_API_KEY=sk-aaa volumes: open-terminal: open-webui: ``` 2. Verify that owui reaches open-terminal: `docker exec -it open-webui curl http://open-terminal:8000/docs` 3. Create login, add OpenAI-API-Key 4. Configure Tool: <img width="480" height="708" alt="Image" src="https://github.com/user-attachments/assets/82991fd4-684c-48ac-81cf-011110ab393f" /> 5. Add Tool to Chat with native mode: <img width="1439" height="622" alt="Image" src="https://github.com/user-attachments/assets/87559a1b-d4e1-4f3c-8ca6-8c66aae083a7" />

bsakaguchi commented

Author

Owner

@silenceroom commented on GitHub (Feb 26, 2026):

@silenceroom @Seviks @MarceloMassarente we're unable to reproduce with the lastest dev with gpt-5.2 (native function calling enabled)

@tjbck I've deployed the dev branch with owui and latest open-terminal, with the gpt-5.2@Azure AI, the problem remains.

@silenceroom commented on GitHub (Feb 26, 2026): > [@silenceroom](https://github.com/silenceroom) [@Seviks](https://github.com/Seviks) [@MarceloMassarente](https://github.com/MarceloMassarente) we're unable to reproduce with the lastest dev with gpt-5.2 (native function calling enabled) @tjbck I've deployed the dev branch with owui and latest open-terminal, with the gpt-5.2@Azure AI, the problem remains.

bsakaguchi commented

Author

Owner

@cloudtuotuo commented on GitHub (Feb 27, 2026):

I suspect this issue comes from OpenWebUI.

When I use Claude-4.6 via New-API to perform Notes-related actions (Builtin Tools), I’ve noticed a consistent failure pattern: if the assistant outputs no textual content at all and only executes a tool call, then the next tool invocation fails with a 400 error. The log looks like this:

record error log: userId=1, channelId=7, modelName=claude-4.6, tokenName=OpenWebUI,
content=status_code=400, messages: text content blocks must be non-empty

I asked Gemini about this error. It suggested that Claude does not accept empty text content blocks, which matches the error message (“text content blocks must be non-empty”). As a workaround, if I need Claude to use tools, I should prompt the assistant to always output at least some text (even a short confirmation) before each tool call, instead of returning a tool-only response.

@cloudtuotuo commented on GitHub (Feb 27, 2026): I suspect this issue comes from OpenWebUI. When I use Claude-4.6 via New-API to perform Notes-related actions (Builtin Tools), I’ve noticed a consistent failure pattern: if the assistant outputs **no textual content at all** and only executes a tool call, then the **next** tool invocation fails with a 400 error. The log looks like this: ``` record error log: userId=1, channelId=7, modelName=claude-4.6, tokenName=OpenWebUI, content=status_code=400, messages: text content blocks must be non-empty ``` I asked Gemini about this error. It suggested that **Claude does not accept empty text content blocks**, which matches the error message (“text content blocks must be non-empty”). As a workaround, if I need Claude to use tools, I should prompt the assistant to **always output at least some text** (even a short confirmation) **before each tool call**, instead of returning a tool-only response.

bsakaguchi commented

Author

Owner

@tjbck commented on GitHub (Feb 27, 2026):

@silenceroom @Seviks @MarceloMassarente could anyone confirm again (latest dev) with a dedicated open terminal integration section?

@tjbck commented on GitHub (Feb 27, 2026): @silenceroom @Seviks @MarceloMassarente could anyone confirm again (latest dev) with a dedicated open terminal integration section?

bsakaguchi commented

Author

Owner

@ExZyle commented on GitHub (Feb 28, 2026):

@silenceroom @Seviks @MarceloMassarente could anyone confirm again (latest dev) with a dedicated open terminal integration section?

I was using ghcr.io/open-webui/open-webui:main from about a day or two ago with the above error, using ghcr.io/open-webui/open-webui:dev works

@ExZyle commented on GitHub (Feb 28, 2026): > [@silenceroom](https://github.com/silenceroom) [@Seviks](https://github.com/Seviks) [@MarceloMassarente](https://github.com/MarceloMassarente) could anyone confirm again (latest dev) with a dedicated open terminal integration section? I was using `ghcr.io/open-webui/open-webui:main` from about a day or two ago with the above error, using `ghcr.io/open-webui/open-webui:dev` works

bsakaguchi commented

Author

Owner

@silenceroom commented on GitHub (Feb 28, 2026):

@silenceroom @Seviks @MarceloMassarente could anyone confirm again (latest dev) with a dedicated open terminal integration section?

@tjbck I can confirm with the latest dev branch the problem now disapear. gpt-5/gpt-5.2(via Azure) now gets response.

@silenceroom commented on GitHub (Feb 28, 2026): > [@silenceroom](https://github.com/silenceroom) [@Seviks](https://github.com/Seviks) [@MarceloMassarente](https://github.com/MarceloMassarente) could anyone confirm again (latest dev) with a dedicated open terminal integration section? @tjbck I can confirm with the latest dev branch the problem now disapear. gpt-5/gpt-5.2(via Azure) now gets response.

bsakaguchi commented