github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 16:11:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #14614] /set think {high|medium|low|true|false} in the TUI crashes qwen3.5 models #9470

New Issue
Open
opened 2026-04-12 22:23:50 -05:00 by GiteaMirror · 4 comments
GiteaMirror commented 2026-04-12 22:23:50 -05:00
Owner
Copy Link

Originally created by @sammyf on GitHub (Mar 4, 2026).
Original GitHub issue: https://github.com/ollama/ollama/issues/14614

What is the issue?

setting think to any of the accepted values returns first an error and eventually crashes the application with a 400 Bad Request

>>> /set think false
Set 'think' mode to 'false'.
>>> hey there
error: 400 Bad Request: invalid think value: "false" (must be "high", "medium", "low", true, or false)
>>> /set think high
Set 'think' mode to 'high'.
>>> hey there
Error: 400 Bad Request: think value "high" is not supported for this model
10:47:16 sammy@neo-bandito ~$

this seems to happen with all official versions of qwen3.5

Relevant log output

there are no relevant logs, as the error seems to happen in the TUI

 
Mar 04 10:47:15 neo-bandito ollama[1594]: [GIN] 2026/03/04 - 10:47:15 | 400 |  146.015127ms |       127.0.0.1 | POST     "/api/chat"
Mar 04 10:47:15 neo-bandito ollama[1594]: [GIN] 2026/03/04 - 10:47:15 | 400 |  146.015127ms |       127.0.0.1 | POST     "/api/chat"

OS

Linux

GPU

Nvidia

CPU

Intel

Ollama version

ollama version is 0.17.5

Originally created by @sammyf on GitHub (Mar 4, 2026). Original GitHub issue: https://github.com/ollama/ollama/issues/14614 ### What is the issue? setting `think` to any of the accepted values returns first an error and eventually crashes the application with a 400 Bad Request ``` >>> /set think false Set 'think' mode to 'false'. >>> hey there error: 400 Bad Request: invalid think value: "false" (must be "high", "medium", "low", true, or false) >>> /set think high Set 'think' mode to 'high'. >>> hey there Error: 400 Bad Request: think value "high" is not supported for this model 10:47:16 sammy@neo-bandito ~$ ``` this seems to happen with all official versions of qwen3.5 ### Relevant log output ```shell there are no relevant logs, as the error seems to happen in the TUI Mar 04 10:47:15 neo-bandito ollama[1594]: [GIN] 2026/03/04 - 10:47:15 | 400 | 146.015127ms | 127.0.0.1 | POST "/api/chat" Mar 04 10:47:15 neo-bandito ollama[1594]: [GIN] 2026/03/04 - 10:47:15 | 400 | 146.015127ms | 127.0.0.1 | POST "/api/chat" ``` ### OS Linux ### GPU Nvidia ### CPU Intel ### Ollama version ollama version is 0.17.5
GiteaMirror added the bug label 2026-04-12 22:23:50 -05:00
GiteaMirror commented 2026-04-12 22:23:52 -05:00
Author
Owner
Copy Link

@fabianfiorotto commented on GitHub (Mar 4, 2026):

The commands for enabling and disabling thinking are

Enable thinking
/set think

Disable thinking
/set nothink

<!-- gh-comment-id:3999646799 --> @fabianfiorotto commented on GitHub (Mar 4, 2026): The commands for enabling and disabling thinking are **Enable thinking** /set think **Disable thinking** /set nothink
GiteaMirror commented 2026-04-12 22:23:54 -05:00
Author
Owner
Copy Link

@lennarkivistik commented on GitHub (Mar 5, 2026):

The thinking parameter from the api has worked great for me but i have only seen
true and false as the only viable options or atleast that is how I interpret the huggingface docs at Qwen3.5:9b

so "high", "medium", "low" is most probably only usefull for gpt-oss models

<!-- gh-comment-id:4003534582 --> @lennarkivistik commented on GitHub (Mar 5, 2026): The thinking parameter from the api has worked great for me but i have only seen true and false as the only viable options or atleast that is how I interpret the huggingface docs at [Qwen3.5:9b](https://huggingface.co/Qwen/Qwen3.5-9B#instruct-or-non-thinking-mode) so "high", "medium", "low" is most probably only usefull for **gpt-oss** models
GiteaMirror commented 2026-04-12 22:23:54 -05:00
Author
Owner
Copy Link

@liaoweiguo commented on GitHub (Mar 8, 2026):

how to set think low in curl

<!-- gh-comment-id:4018298332 --> @liaoweiguo commented on GitHub (Mar 8, 2026): how to set think low in curl
GiteaMirror commented 2026-04-12 22:23:55 -05:00
Author
Owner
Copy Link

@lennarkivistik commented on GitHub (Mar 8, 2026):

for Ollama, set think to "low" for gpt-oss:20b. Ollama’s docs say GPT-OSS expects think to be one of "low", "medium", or "high" rather than true/false.

curl http://localhost:11434/api/chat \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-oss:20b",
    "messages": [
      {
        "role": "user",
        "content": "Write a short product description for a compact mechanical keyboard."
      }
    ],
    "think": "low",
    "stream": false
  }'

Whilst qwen3.5 expects true and false

<!-- gh-comment-id:4018525435 --> @lennarkivistik commented on GitHub (Mar 8, 2026): for Ollama, set think to "low" for gpt-oss:20b. Ollama’s docs say GPT-OSS expects think to be one of "low", "medium", or "high" rather than true/false. ```` curl http://localhost:11434/api/chat \ -H "Content-Type: application/json" \ -d '{ "model": "gpt-oss:20b", "messages": [ { "role": "user", "content": "Write a short product description for a compact mechanical keyboard." } ], "think": "low", "stream": false }' ```` Whilst qwen3.5 expects true and false
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label bug
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#9470
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?