I have searched for all existing open AND closed issues and discussions for similar requests. I have found none that is comparable to my request.
Verify Feature Scope
I have read through and understood the scope definition for feature requests in the Issues section. I believe my feature request meets the definition and belongs in the Issues section instead of the Discussions.
Problem Description
No support for Ollama image generator
Desired Solution you'd like
Ollama has Added experimental MLX backend and engine with imagegen support.
OpenAPI option should work when this is released, however I think it should have it's own option using ollama settings already listed in the web UI.
Originally created by @Hello-World-Traveler on GitHub (Jan 10, 2026).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/20543
### Check Existing Issues
- [x] I have searched for all existing **open AND closed** issues and discussions for similar requests. I have found none that is comparable to my request.
### Verify Feature Scope
- [x] I have read through and understood the scope definition for feature requests in the Issues section. I believe my feature request meets the definition and belongs in the Issues section instead of the Discussions.
### Problem Description
No support for Ollama image generator
### Desired Solution you'd like
Ollama has Added experimental MLX backend and engine with imagegen support.
<img width="1409" height="138" alt="Image" src="https://github.com/user-attachments/assets/e4873a1b-579f-42d7-9898-1edac51caf1c" />
OpenAPI option should work when this is released, however I think it should have it's own option using ollama settings already listed in the web UI.
https://github.com/ollama/ollama/releases/tag/v0.14.0-rc2
### Alternatives Considered
_No response_
### Additional Context
_No response_
Just my two cents: Let’s wait for this feature to stabilize on Ollama’s side and see how things pan out from there. For all we know (and we don't know), this feature could be stripped out and scrapped before the next official version release of Ollama. It's currently an experimental feature, likely macOS-only out the gate, and appears only in a pre-release changelog. We don't want to jump the gun and prematurely build support for something which might not ship.
<!-- gh-comment-id:3733531697 -->
@silentoplayz commented on GitHub (Jan 10, 2026):
Just my two cents: Let’s wait for this feature to stabilize on Ollama’s side and see how things pan out from there. For all we know (and we don't know), this feature could be stripped out and scrapped before the next official version release of Ollama. It's currently an experimental feature, likely macOS-only out the gate, and appears only in a pre-release changelog. We don't want to jump the gun and prematurely build support for something which might not ship.
@Hello-World-Traveler commented on GitHub (Jan 10, 2026):
I agree. This issue was to make open-webui team aware of this upcoming feature.
<!-- gh-comment-id:3733549881 -->
@Hello-World-Traveler commented on GitHub (Jan 10, 2026):
I agree. This issue was to make open-webui team aware of this upcoming feature.
The PR above is still experimental (x/imagegen, MLX runner, Z-Image only, no Ollama-native API) and explicitly “not optimized.”
Open WebUI already speaks the OpenAI /v1/images/generations endpoint natively as far as I'm aware, so nothing needs to be added.
<!-- gh-comment-id:3733552450 -->
@silentoplayz commented on GitHub (Jan 10, 2026):
Thank you for bringing this upcoming feature in Ollama to our attention.
I will be closing this issue for now. We’ll wait for Ollama to stabilize its image-generation path. For reference, see https://github.com/ollama/ollama/pull/13659.
The PR above is still experimental (x/imagegen, MLX runner, Z-Image only, no Ollama-native API) and explicitly “not optimized.”
Open WebUI already speaks the OpenAI `/v1/images/generations` endpoint natively as far as I'm aware, so nothing needs to be added.
@silentoplayz the current /v1/images/generations is not to fit the openai endpoint cause it only resposes with a string so they are sadly not equal
<!-- gh-comment-id:3744028937 -->
@Keugon commented on GitHub (Jan 13, 2026):
@silentoplayz the current /v1/images/generations is not to fit the openai endpoint cause it only resposes with a string so they are sadly not equal
@Hello-World-Traveler commented on GitHub (Jan 18, 2026):
Next update should have:
API changes:
Add width, height, steps parameters to GenerateRequest
Add status, total, completed, images fields to GenerateResponse
Return base64-encoded image data in images array when done
OpenAI compatibility:
Add /v1/images/generations endpoint using middleware pattern
Transform OpenAI image generation requests to /api/generate format
Return OpenAI-compatible response format with b64_json data
<!-- gh-comment-id:3765753615 -->
@Hello-World-Traveler commented on GitHub (Jan 18, 2026):
Next update should have:
API changes:
Add width, height, steps parameters to GenerateRequest
Add status, total, completed, images fields to GenerateResponse
Return base64-encoded image data in images array when done
OpenAI compatibility:
Add /v1/images/generations endpoint using middleware pattern
Transform OpenAI image generation requests to /api/generate format
Return OpenAI-compatible response format with b64_json data
Apparently Ollama is extending their Ollama API:
From the changelog: "The /api/generate API can now be used for image generation"
Open WebUI probably neet another Connection type in Image Generation that is Ollama - similar how Ollama is a Connection type already on the normal connections in admin panel.
OR MAYBE Ollama will not need to be connected like an typical Image Generation API but just how it currently is with the Connections. But then Open WebUI probably needs bigger changes to handle the new logic?
<!-- gh-comment-id:3780569953 -->
@TomTheWise commented on GitHub (Jan 21, 2026):
https://github.com/ollama/ollama/releases/tag/v0.14.3-rc3
Apparently Ollama is extending their Ollama API:
From the changelog: "The /api/generate API can now be used for image generation"
Open WebUI probably neet another Connection type in Image Generation that is Ollama - similar how Ollama is a Connection type already on the normal connections in admin panel.
OR MAYBE Ollama will not need to be connected like an typical Image Generation API but just how it currently is with the Connections. But then Open WebUI probably needs bigger changes to handle the new logic?
Image generation in Ollama only works on MacOS for the time being.
<!-- gh-comment-id:3821100635 -->
@Hello-World-Traveler commented on GitHub (Jan 30, 2026):
Image generation documentation Draft
https://github.com/ollama/ollama/pull/13849/commits/8b4410633d2f749e0b2fa8da182b29763a94f511
https://platform.openai.com/docs/api-reference/images
Image generation in Ollama only works on MacOS for the time being.
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 @Hello-World-Traveler on GitHub (Jan 10, 2026).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/20543
Check Existing Issues
Verify Feature Scope
Problem Description
No support for Ollama image generator
Desired Solution you'd like
Ollama has Added experimental MLX backend and engine with imagegen support.
OpenAPI option should work when this is released, however I think it should have it's own option using ollama settings already listed in the web UI.
https://github.com/ollama/ollama/releases/tag/v0.14.0-rc2
Alternatives Considered
No response
Additional Context
No response
@silentoplayz commented on GitHub (Jan 10, 2026):
Just my two cents: Let’s wait for this feature to stabilize on Ollama’s side and see how things pan out from there. For all we know (and we don't know), this feature could be stripped out and scrapped before the next official version release of Ollama. It's currently an experimental feature, likely macOS-only out the gate, and appears only in a pre-release changelog. We don't want to jump the gun and prematurely build support for something which might not ship.
@Hello-World-Traveler commented on GitHub (Jan 10, 2026):
I agree. This issue was to make open-webui team aware of this upcoming feature.
@silentoplayz commented on GitHub (Jan 10, 2026):
Thank you for bringing this upcoming feature in Ollama to our attention.
I will be closing this issue for now. We’ll wait for Ollama to stabilize its image-generation path. For reference, see https://github.com/ollama/ollama/pull/13659.
The PR above is still experimental (x/imagegen, MLX runner, Z-Image only, no Ollama-native API) and explicitly “not optimized.”
Open WebUI already speaks the OpenAI
/v1/images/generationsendpoint natively as far as I'm aware, so nothing needs to be added.@Keugon commented on GitHub (Jan 13, 2026):
@silentoplayz the current /v1/images/generations is not to fit the openai endpoint cause it only resposes with a string so they are sadly not equal
@Hello-World-Traveler commented on GitHub (Jan 18, 2026):
Next update should have:
API changes:
OpenAI compatibility:
@TomTheWise commented on GitHub (Jan 21, 2026):
https://github.com/ollama/ollama/releases/tag/v0.14.3-rc3
Apparently Ollama is extending their Ollama API:
From the changelog: "The /api/generate API can now be used for image generation"
Open WebUI probably neet another Connection type in Image Generation that is Ollama - similar how Ollama is a Connection type already on the normal connections in admin panel.
OR MAYBE Ollama will not need to be connected like an typical Image Generation API but just how it currently is with the Connections. But then Open WebUI probably needs bigger changes to handle the new logic?
@Hello-World-Traveler commented on GitHub (Jan 30, 2026):
Image generation documentation Draft
https://github.com/ollama/ollama/pull/13849/commits/8b4410633d2f749e0b2fa8da182b29763a94f511
https://platform.openai.com/docs/api-reference/images
Image generation in Ollama only works on MacOS for the time being.
@TomTheWise commented on GitHub (Jan 30, 2026):
Yes but the api is already in place and its only a Matter of time.
@tjbck commented on GitHub (Jan 31, 2026):
https://docs.ollama.com/api/openai-compatibility#%2Fv1%2Fimages%2Fgenerations-experimental