issue: /api/chat/completions in streaming mode set "finish_reason": "stop" for reasoning_content chunk #5991

New Issue

2025-11-11T16:41:46-06:00

GiteaMirror commented

2025-11-11 16:41:46 -06:00

Originally created by @SuperPat45 on GitHub (Aug 8, 2025).

Check Existing Issues

I have searched the existing issues and discussions.
I am using the latest version of Open WebUI.

Installation Method

Docker

Open WebUI Version

v0.6.18

Ollama Version (if applicable)

v0.11.4

Operating System

Linux

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

"finish_reason" should be null for reasoning_content chunk

Actual Behavior

"finish_reason" is always "stop" for reasoning_content chunk

Steps to Reproduce

Configure the gpt-oss:latest model with ollama
Call the /api/chat/completions API model in streaming mode:

{
  "messages":[{
    "content":"...",
    "role":"user"
  }],
  "model":"gpt-oss:latest",
  "stream":true
}

All reasoning_content chunks have "finish_reason": "stop" instead of "finish_reason": null.
This can break third party applications which relies on this parameter to stop the display.

Logs & Screenshots

data: {"id": "gpt-oss:latest-eba6d37d-3187-421f-a904-5a2f1b633b8f", "created": 1754645524, "model": "gpt-oss:latest", "choices": [{"index": 0, "logprobs": null, "finish_reason": "stop", "delta": {"reasoning_content": "We"}}], "object": "chat.completion.chunk"}
data: {"id": "gpt-oss:latest-e8a0654a-a490-411f-9304-fd060511a906", "created": 1754645524, "model": "gpt-oss:latest", "choices": [{"index": 0, "logprobs": null, "finish_reason": "stop", "delta": {"reasoning_content": " need"}}], "object": "chat.completion.chunk"}
data: {"id": "gpt-oss:latest-4240a398-d121-4da3-9e63-d150700edec4", "created": 1754645524, "model": "gpt-oss:latest", "choices": [{"index": 0, "logprobs": null, "finish_reason": "stop", "delta": {"reasoning_content": " answer"}}], "object": "chat.completion.chunk"}
....

Additional Information

No response

Originally created by @SuperPat45 on GitHub (Aug 8, 2025). ### Check Existing Issues - [x] I have searched the existing issues and discussions. - [x] I am using the latest version of Open WebUI. ### Installation Method Docker ### Open WebUI Version v0.6.18 ### Ollama Version (if applicable) v0.11.4 ### Operating System Linux ### 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 "finish_reason" should be null for reasoning_content chunk ### Actual Behavior "finish_reason" is always "stop" for reasoning_content chunk ### Steps to Reproduce Configure the gpt-oss:latest model with ollama Call the /api/chat/completions API model in streaming mode: ``` { "messages":[{ "content":"...", "role":"user" }], "model":"gpt-oss:latest", "stream":true } ``` All reasoning_content chunks have "finish_reason": "stop" instead of "finish_reason": null. This can break third party applications which relies on this parameter to stop the display. ### Logs & Screenshots ``` data: {"id": "gpt-oss:latest-eba6d37d-3187-421f-a904-5a2f1b633b8f", "created": 1754645524, "model": "gpt-oss:latest", "choices": [{"index": 0, "logprobs": null, "finish_reason": "stop", "delta": {"reasoning_content": "We"}}], "object": "chat.completion.chunk"} data: {"id": "gpt-oss:latest-e8a0654a-a490-411f-9304-fd060511a906", "created": 1754645524, "model": "gpt-oss:latest", "choices": [{"index": 0, "logprobs": null, "finish_reason": "stop", "delta": {"reasoning_content": " need"}}], "object": "chat.completion.chunk"} data: {"id": "gpt-oss:latest-4240a398-d121-4da3-9e63-d150700edec4", "created": 1754645524, "model": "gpt-oss:latest", "choices": [{"index": 0, "logprobs": null, "finish_reason": "stop", "delta": {"reasoning_content": " answer"}}], "object": "chat.completion.chunk"} .... ``` ### Additional Information _No response_

GiteaMirror added the bug label 2025-11-11 16:41:46 -06:00

GiteaMirror closed this issue

2025-11-11 16:41:46 -06:00

GiteaMirror commented

2025-11-11 16:41:47 -06:00