github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 08:30:05 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #13468] Vision capable cloud models do not have an Image input mode in the library. #70946

New Issue
Closed
opened 2026-05-04 23:32:30 -05:00 by GiteaMirror · 3 comments
GiteaMirror commented 2026-05-04 23:32:30 -05:00
Owner
Copy Link

Originally created by @rick-github on GitHub (Dec 14, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/13468

What is the issue?

These Cloud models are vision capable:

qwen3-vl:235b-instruct
qwen3-vl:235b
ministral-3:3b
ministral-3:8b
ministral-3:14b
mistral-large-3:675b
devstral-small-2:24b
gemini-3-pro-preview
gemma3:4b
gemma3:12b
gemma3:27b

However, the listings in the library show only Text as an input mode:

Image

Relevant log output

OS

No response

GPU

No response

CPU

No response

Ollama version

No response

Originally created by @rick-github on GitHub (Dec 14, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/13468 ### What is the issue? These Cloud models are vision capable: qwen3-vl:235b-instruct qwen3-vl:235b ministral-3:3b ministral-3:8b ministral-3:14b mistral-large-3:675b devstral-small-2:24b gemini-3-pro-preview gemma3:4b gemma3:12b gemma3:27b However, the listings in the library show only `Text` as an input mode: <img width="802" height="676" alt="Image" src="https://github.com/user-attachments/assets/32af5554-7fa6-43d7-9d74-e22532ca8ac0" /> ### Relevant log output ```shell ``` ### OS _No response_ ### GPU _No response_ ### CPU _No response_ ### Ollama version _No response_
GiteaMirror added the cloudbugollama.com labels 2026-05-04 23:32:32 -05:00
GiteaMirror commented 2026-05-04 23:32:43 -05:00
Author
Owner
Copy Link

@para-droid-ai commented on GitHub (Dec 25, 2025):

Image

gemini 3 flash-cloud
Not sure "what" is happening but I do recall this workflow "working" previously. However, I don't recall if it was a local model or cloud, so I can check that and report back. I don't heavily use the ollama app, mostly just to fire things off at specific open source models.

Ollama Version:

Ollama 0.13.5

<!-- gh-comment-id:3691013629 --> @para-droid-ai commented on GitHub (Dec 25, 2025): <img width="829" height="348" alt="Image" src="https://github.com/user-attachments/assets/28eeb279-6665-455f-8fcb-7892d19a775e" /> **gemini 3 flash-cloud** Not sure "what" is happening but I do recall this workflow "working" previously. However, I don't recall if it was a local model or cloud, so I can check that and report back. I don't heavily use the ollama app, mostly just to fire things off at specific open source models. ### Ollama Version: Ollama 0.13.5
GiteaMirror commented 2026-05-04 23:32:46 -05:00
Author
Owner
Copy Link

@para-droid-ai commented on GitHub (Dec 25, 2025):

revised with Gemini and some digging:

CRITICAL: Systemic UI Regression on v0.13.5 (Gemini 3 + Qwen 3 VL)

This is not isolated to cloud proxies. The Desktop App is aggressively disabling Vision input for multiple new major releases:

  1. gemini-3-flash-preview:cloud (Vision/Tools/Thinking)
  2. qwen3-vl (Native Vision-Language Model)

Behavior:
Despite both models having vision tags in the library, the App UI hides the image attachment button and throws "This model does not support images" on paste.

Diagnosis:
The client manifest parser seems to choke on newer capability sets (e.g., models that combine vision + thinking or tools), failing safe to "Text Only."

Reproduction:
ollama pull qwen3-vl -> Open App -> No Image Button.

<!-- gh-comment-id:3691063028 --> @para-droid-ai commented on GitHub (Dec 25, 2025): revised with Gemini and some digging: **CRITICAL: Systemic UI Regression on v0.13.5 (Gemini 3 + Qwen 3 VL)** This is not isolated to cloud proxies. The Desktop App is aggressively disabling Vision input for multiple new major releases: 1. `gemini-3-flash-preview:cloud` (Vision/Tools/Thinking) 2. `qwen3-vl` (Native Vision-Language Model) **Behavior:** Despite both models having `vision` tags in the library, the App UI hides the image attachment button and throws "This model does not support images" on paste. **Diagnosis:** The client manifest parser seems to choke on newer capability sets (e.g., models that combine `vision` + `thinking` or `tools`), failing safe to "Text Only." **Reproduction:** `ollama pull qwen3-vl` -> Open App -> No Image Button.
GiteaMirror commented 2026-05-04 23:32:48 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Dec 25, 2025):

CRITICAL: Systemic UI Regression on v0.13.5 (Gemini 3 + Qwen 3 VL)

Not related. See #13459.

<!-- gh-comment-id:3691214970 --> @rick-github commented on GitHub (Dec 25, 2025): > CRITICAL: Systemic UI Regression on v0.13.5 (Gemini 3 + Qwen 3 VL) Not related. See #13459.
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 cloud ollama.com
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#70946
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?