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 #6748] Support Mistral's new visual model: Pixtral-12b-240910 #30013

New Issue
Closed
opened 2026-04-22 09:25:03 -05:00 by GiteaMirror · 52 comments
GiteaMirror commented 2026-04-22 09:25:03 -05:00
Owner
Copy Link

Originally created by @awaescher on GitHub (Sep 11, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/6748

Mistral AI just dropped Pixtral, their 12b model with vision support.

Originally created by @awaescher on GitHub (Sep 11, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/6748 Mistral AI just dropped Pixtral, their 12b model with vision support. - https://github.com/mistralai/mistral-common/releases/tag/v1.4.0 - https://www.reddit.com/r/LocalLLaMA/comments/1fe3x1z/mistral_dropping_a_new_magnet_link/
GiteaMirror added the model label 2026-04-22 09:25:03 -05:00
GiteaMirror commented 2026-04-22 09:25:04 -05:00
Author
Owner
Copy Link

@Qualzz commented on GitHub (Sep 11, 2024):

Gosh I love them.

<!-- gh-comment-id:2343159080 --> @Qualzz commented on GitHub (Sep 11, 2024): Gosh I love them.
GiteaMirror commented 2026-04-22 09:25:04 -05:00
Author
Owner
Copy Link

@Kerimcanak commented on GitHub (Sep 11, 2024):

Support.
I want to see this model to be usable on Ollama.

<!-- gh-comment-id:2343948271 --> @Kerimcanak commented on GitHub (Sep 11, 2024): Support. I want to see this model to be usable on Ollama.
GiteaMirror commented 2026-04-22 09:25:04 -05:00
Author
Owner
Copy Link

@tuanlda78202 commented on GitHub (Sep 12, 2024):

Hope too ~

<!-- gh-comment-id:2345652587 --> @tuanlda78202 commented on GitHub (Sep 12, 2024): Hope too ~
GiteaMirror commented 2026-04-22 09:25:05 -05:00
Author
Owner
Copy Link

@briansan commented on GitHub (Sep 12, 2024):

+1

<!-- gh-comment-id:2346859846 --> @briansan commented on GitHub (Sep 12, 2024): +1
GiteaMirror commented 2026-04-22 09:25:05 -05:00
Author
Owner
Copy Link

@awaescher commented on GitHub (Sep 13, 2024):

Kudos to the Ollama team ❤️

image

(source)

<!-- gh-comment-id:2349045484 --> @awaescher commented on GitHub (Sep 13, 2024): Kudos to the Ollama team :heart: ![image](https://github.com/user-attachments/assets/6b713144-538b-4f11-9b76-d41f9fa4d72e) [(source)](https://github.com/ollama/ollama/issues/6776#issuecomment-2347312012)
GiteaMirror commented 2026-04-22 09:25:06 -05:00
Author
Owner
Copy Link

@abenmrad commented on GitHub (Sep 16, 2024):

Related to Pixtral, but more generally regarding Multimodal support in Ollama:

From my experiments today, Ollama is still supporting multi-modal chat with LLaVa (retried today with v0.3.10).

There were indeed some changes in the Llama.cpp server a while back.

I was genuinely interested to understand how Ollama can still handle it while Llama.cpp reportedly cannot anymore. Was Ollama relying on llama-cli, not llama-server?

Turns out it's not relying on either. Ollama is integrating directly with the llama.cpp code base. As for LLaVa support, they lifted the LLaVa support directly from the Llama.cpp server codebase and have been maintaining that in addition to everything else since then...

Ollama team are truly unsung heroes in this technological revolution.

<!-- gh-comment-id:2353392349 --> @abenmrad commented on GitHub (Sep 16, 2024): Related to Pixtral, but more generally regarding Multimodal support in Ollama: From my experiments today, Ollama is still supporting multi-modal chat with LLaVa (retried today with v0.3.10). There were indeed [some changes in the Llama.cpp server](https://github.com/ggerganov/llama.cpp/pull/5882) a while back. I was genuinely interested to understand how Ollama can still handle it while Llama.cpp reportedly cannot anymore. Was Ollama relying on `llama-cli`, not `llama-server`? Turns out it's not relying on either. Ollama is [integrating directly with the llama.cpp code base](https://github.com/ollama/ollama/blob/main/llm/llm.go). As for LLaVa support, they [lifted the LLaVa support directly from the Llama.cpp server codebase](https://github.com/ollama/ollama/pull/3086) and have been maintaining that in addition to everything else since then... Ollama team are truly unsung heroes in this technological revolution.
GiteaMirror commented 2026-04-22 09:25:06 -05:00
Author
Owner
Copy Link

@BennisonDevadoss commented on GitHub (Sep 19, 2024):

Is it available at Ollama now?

<!-- gh-comment-id:2359988087 --> @BennisonDevadoss commented on GitHub (Sep 19, 2024): Is it available at Ollama now?
GiteaMirror commented 2026-04-22 09:25:07 -05:00
Author
Owner
Copy Link

@webboty commented on GitHub (Sep 19, 2024):

Is it available at Ollama now?

ollama pull pixtral
pulling manifest
Error: pull model manifest: file does not exist

Not yet.

<!-- gh-comment-id:2360840418 --> @webboty commented on GitHub (Sep 19, 2024): > Is it available at Ollama now? ollama pull pixtral pulling manifest Error: pull model manifest: file does not exist Not yet.
GiteaMirror commented 2026-04-22 09:25:07 -05:00
Author
Owner
Copy Link

@dropout-developer commented on GitHub (Sep 19, 2024):

Is it available at Ollama now?

ollama pull pixtral pulling manifest Error: pull model manifest: file does not exist

Not yet.

Same for me

<!-- gh-comment-id:2361849738 --> @dropout-developer commented on GitHub (Sep 19, 2024): > > Is it available at Ollama now? > > ollama pull pixtral pulling manifest Error: pull model manifest: file does not exist > > Not yet. Same for me
GiteaMirror commented 2026-04-22 09:25:08 -05:00
Author
Owner
Copy Link

@CRCODE22 commented on GitHub (Sep 19, 2024):

Tried to convert and add this with ollama https://huggingface.co/DewEfresh/pixtral-12b-8bit/tree/main but it seems the architecture is not supported by Ollama (yet).

K:\AI\DewEfresh\pixtral-12b-8bit>ollama create Pixtral-12B-2409
transferring model data 100%
converting model
Error: unsupported architecture

K:\AI\DewEfresh\pixtral-12b-8bit>ollama create --quantize q8_K_M Pixtral-12B-2409
transferring model data 100%
converting model
Error: unsupported architecture

<!-- gh-comment-id:2361948020 --> @CRCODE22 commented on GitHub (Sep 19, 2024): Tried to convert and add this with ollama https://huggingface.co/DewEfresh/pixtral-12b-8bit/tree/main but it seems the architecture is not supported by Ollama (yet). K:\AI\DewEfresh\pixtral-12b-8bit>ollama create Pixtral-12B-2409 transferring model data 100% converting model Error: unsupported architecture K:\AI\DewEfresh\pixtral-12b-8bit>ollama create --quantize q8_K_M Pixtral-12B-2409 transferring model data 100% converting model Error: unsupported architecture
GiteaMirror commented 2026-04-22 09:25:08 -05:00
Author
Owner
Copy Link

@celobusana commented on GitHub (Sep 19, 2024):

I tried to run the Pixtral with python code using an RTX 4060 with 16GB, but it was not possible :(. Perhaps it would work with a 4090 with 24GB.

torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 60.00 MiB. GPU 0 has a total capacity of 15.70 GiB of which 34.69 MiB is free. Including non-PyTorch memory, this process has 15.64 GiB memory in use. Of the allocated memory 15.50 GiB is allocated by PyTorch, and 10.20 MiB is reserved by PyTorch but unallocated.

<!-- gh-comment-id:2362142668 --> @celobusana commented on GitHub (Sep 19, 2024): I tried to run the Pixtral with python code using an RTX 4060 with 16GB, but it was not possible :(. Perhaps it would work with a 4090 with 24GB. `torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 60.00 MiB. GPU 0 has a total capacity of 15.70 GiB of which 34.69 MiB is free. Including non-PyTorch memory, this process has 15.64 GiB memory in use. Of the allocated memory 15.50 GiB is allocated by PyTorch, and 10.20 MiB is reserved by PyTorch but unallocated.`
GiteaMirror commented 2026-04-22 09:25:08 -05:00
Author
Owner
Copy Link

@thecr7guy2 commented on GitHub (Sep 20, 2024):

Cant wait to try the gguf version of Pixtral man
Ollama gonna deliver a christmas present early for us

<!-- gh-comment-id:2363687339 --> @thecr7guy2 commented on GitHub (Sep 20, 2024): Cant wait to try the gguf version of Pixtral man Ollama gonna deliver a christmas present early for us
GiteaMirror commented 2026-04-22 09:25:09 -05:00
Author
Owner
Copy Link

@ddpasa commented on GitHub (Sep 22, 2024):

I tried to run the Pixtral with python code using an RTX 4060 with 16GB, but it was not possible :(. Perhaps it would work with a 4090 with 24GB.

torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 60.00 MiB. GPU 0 has a total capacity of 15.70 GiB of which 34.69 MiB is free. Including non-PyTorch memory, this process has 15.64 GiB memory in use. Of the allocated memory 15.50 GiB is allocated by PyTorch, and 10.20 MiB is reserved by PyTorch but unallocated.

You need to quantize to run an 12b model on 16GB hardware.

<!-- gh-comment-id:2366786318 --> @ddpasa commented on GitHub (Sep 22, 2024): > I tried to run the Pixtral with python code using an RTX 4060 with 16GB, but it was not possible :(. Perhaps it would work with a 4090 with 24GB. > > `torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 60.00 MiB. GPU 0 has a total capacity of 15.70 GiB of which 34.69 MiB is free. Including non-PyTorch memory, this process has 15.64 GiB memory in use. Of the allocated memory 15.50 GiB is allocated by PyTorch, and 10.20 MiB is reserved by PyTorch but unallocated.` You need to quantize to run an 12b model on 16GB hardware.
GiteaMirror commented 2026-04-22 09:25:09 -05:00
Author
Owner
Copy Link

@Qualzz commented on GitHub (Sep 25, 2024):

any news ?

<!-- gh-comment-id:2375351935 --> @Qualzz commented on GitHub (Sep 25, 2024): any news ?
GiteaMirror commented 2026-04-22 09:25:09 -05:00
Author
Owner
Copy Link

@ep5000 commented on GitHub (Sep 27, 2024):

+1 for this feature

<!-- gh-comment-id:2379163564 --> @ep5000 commented on GitHub (Sep 27, 2024): +1 for this feature
GiteaMirror commented 2026-04-22 09:25:10 -05:00
Author
Owner
Copy Link

@A-REDA-EL commented on GitHub (Sep 27, 2024):

any news ?

<!-- gh-comment-id:2379554836 --> @A-REDA-EL commented on GitHub (Sep 27, 2024): any news ?
GiteaMirror commented 2026-04-22 09:25:11 -05:00
Author
Owner
Copy Link

@frostworx commented on GitHub (Sep 27, 2024):

please stop spamming here.

multiple people are subscribed to this issue and patiently wait until it is done.
all of them receive an email for every "+1" - it is annoying and for sure doesn't speed up anything.
thanks for your understanding.

(sorry for another mail, subscribers...)

<!-- gh-comment-id:2379575418 --> @frostworx commented on GitHub (Sep 27, 2024): please stop spamming here. multiple people are subscribed to this issue and patiently wait until it is done. all of them receive an email for every "+1" - it is annoying and for sure doesn't speed up anything. thanks for your understanding. _(sorry for another mail, subscribers...)_
GiteaMirror commented 2026-04-22 09:25:11 -05:00
Author
Owner
Copy Link

@robert-mcdermott commented on GitHub (Oct 1, 2024):

Anyone else checking the model library a couple times a day waiting for Pixtral, Llama3.2 9B and Molmo-7B to drop? 😄

<!-- gh-comment-id:2387083244 --> @robert-mcdermott commented on GitHub (Oct 1, 2024): Anyone else checking the model library a couple times a day waiting for Pixtral, Llama3.2 9B and Molmo-7B to drop? 😄
GiteaMirror commented 2026-04-22 09:25:12 -05:00
Author
Owner
Copy Link

@Qualzz commented on GitHub (Oct 1, 2024):

im not sure if the team is actively working on multimodal support or if
they're focusing on something else at the moment. What is certain is that
multimodal capabilities will become increasingly essential in the near
future, and many users may switch to alternatives that offer this
functionality.

On Tue, Oct 1, 2024, 23:11 Robert McDermott @.***>
wrote:

Anyone else checking the model library a couple times a day waiting for
Pixtral, Llama3.2 9B and Molmo-7B to drop? 😄

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/6748#issuecomment-2387083244,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AIMKMGFPFQGUUDYTKWQV67TZZMFWZAVCNFSM6AAAAABOANEDJGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGOBXGA4DGMRUGQ
.
You are receiving this because you commented.Message ID:
@.***>

<!-- gh-comment-id:2387262280 --> @Qualzz commented on GitHub (Oct 1, 2024): im not sure if the team is actively working on multimodal support or if they're focusing on something else at the moment. What is certain is that multimodal capabilities will become increasingly essential in the near future, and many users may switch to alternatives that offer this functionality. On Tue, Oct 1, 2024, 23:11 Robert McDermott ***@***.***> wrote: > Anyone else checking the model library a couple times a day waiting for > Pixtral, Llama3.2 9B and Molmo-7B to drop? 😄 > > — > Reply to this email directly, view it on GitHub > <https://github.com/ollama/ollama/issues/6748#issuecomment-2387083244>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AIMKMGFPFQGUUDYTKWQV67TZZMFWZAVCNFSM6AAAAABOANEDJGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGOBXGA4DGMRUGQ> > . > You are receiving this because you commented.Message ID: > ***@***.***> >
GiteaMirror commented 2026-04-22 09:25:12 -05:00
Author
Owner
Copy Link

@sr-tream commented on GitHub (Oct 2, 2024):

in the near future, and many users may switch to alternatives that offer this functionality.

Not with local models, I guess.

<!-- gh-comment-id:2388953748 --> @sr-tream commented on GitHub (Oct 2, 2024): > in the near future, and many users may switch to alternatives that offer this functionality. Not with local models, I guess.
GiteaMirror commented 2026-04-22 09:25:13 -05:00
Author
Owner
Copy Link

@Rakhsan commented on GitHub (Oct 3, 2024):

How much long any hint pls?

<!-- gh-comment-id:2391395692 --> @Rakhsan commented on GitHub (Oct 3, 2024): How much long any hint pls?
GiteaMirror commented 2026-04-22 09:25:13 -05:00
Author
Owner
Copy Link

@wijjj commented on GitHub (Oct 10, 2024):

Anyone else checking the model library a couple times a day waiting for Pixtral, Llama3.2 9B and Molmo-7B to drop? 😄

pretty much

<!-- gh-comment-id:2404787146 --> @wijjj commented on GitHub (Oct 10, 2024): > Anyone else checking the model library a couple times a day waiting for Pixtral, Llama3.2 9B and Molmo-7B to drop? 😄 pretty much
GiteaMirror commented 2026-04-22 09:25:13 -05:00
Author
Owner
Copy Link

@nuryslyrt commented on GitHub (Oct 11, 2024):

+1 to this feature :)

<!-- gh-comment-id:2407498384 --> @nuryslyrt commented on GitHub (Oct 11, 2024): +1 to this feature :)
GiteaMirror commented 2026-04-22 09:25:14 -05:00
Author
Owner
Copy Link

@Rakhsan commented on GitHub (Oct 11, 2024):

Anyone else checking the model library a couple times a day waiting for Pixtral, Llama3.2 9B and Molmo-7B to drop? 😄

pretty much

there is no Llama3.2 9B THERE ARE 1, 3, 11, 90

<!-- gh-comment-id:2407783365 --> @Rakhsan commented on GitHub (Oct 11, 2024): > > Anyone else checking the model library a couple times a day waiting for Pixtral, Llama3.2 9B and Molmo-7B to drop? 😄 > > pretty much there is no Llama3.2 9B THERE ARE 1, 3, 11, 90
GiteaMirror commented 2026-04-22 09:25:14 -05:00
Author
Owner
Copy Link

@wijjj commented on GitHub (Oct 15, 2024):

Anyone else checking the model library a couple times a day waiting for Pixtral, Llama3.2 9B and Molmo-7B to drop? 😄

pretty much

there is no Llama3.2 9B THERE ARE 1, 3, 11, 90

yeah. Thanks. You know what we mean. Anyways: waiting patiently.

<!-- gh-comment-id:2413926361 --> @wijjj commented on GitHub (Oct 15, 2024): > > > Anyone else checking the model library a couple times a day waiting for Pixtral, Llama3.2 9B and Molmo-7B to drop? 😄 > > > > > > pretty much > > there is no Llama3.2 9B THERE ARE 1, 3, 11, 90 yeah. Thanks. You know what we mean. Anyways: waiting patiently.
GiteaMirror commented 2026-04-22 09:25:14 -05:00
Author
Owner
Copy Link

@semidark commented on GitHub (Oct 15, 2024):

please stop spamming here.

multiple people are subscribed to this issue and patiently wait until it is done.
all of them receive an email for every "+1" - it is annoying and for sure doesn't speed up anything.
thanks for your understanding.

(sorry for another mail, subscribers...)

Please do your Daily checks for "the drop" without spamming. Holy flipping cow.

<!-- gh-comment-id:2414833943 --> @semidark commented on GitHub (Oct 15, 2024): > please stop spamming here. > > multiple people are subscribed to this issue and patiently wait until it is done. > all of them receive an email for every "+1" - it is annoying and for sure doesn't speed up anything. > thanks for your understanding. > > _(sorry for another mail, subscribers...)_ > Please do your Daily checks for "the drop" without spamming. Holy flipping cow.
GiteaMirror commented 2026-04-22 09:25:15 -05:00
Author
Owner
Copy Link

@pbasov commented on GitHub (Oct 22, 2024):

Just to point out an alternative for now - LM Studio just released with Pixtral support:
https://lmstudio.ai/blog/lmstudio-v0.3.5

<!-- gh-comment-id:2429746604 --> @pbasov commented on GitHub (Oct 22, 2024): Just to point out an alternative for now - LM Studio just released with Pixtral support: https://lmstudio.ai/blog/lmstudio-v0.3.5
GiteaMirror commented 2026-04-22 09:25:15 -05:00
Author
Owner
Copy Link

@oderwat commented on GitHub (Oct 22, 2024):

@pbasov Do I understand it right that this is Apple-Silicone only?

<!-- gh-comment-id:2430448048 --> @oderwat commented on GitHub (Oct 22, 2024): @pbasov Do I understand it right that this is Apple-Silicone only?
GiteaMirror commented 2026-04-22 09:25:15 -05:00
Author
Owner
Copy Link

@pbasov commented on GitHub (Oct 23, 2024):

@oderwat I believe so, yes, since it's enabled by the MLX engine and llama.cpp still doesn't support it.
https://github.com/ggerganov/llama.cpp/issues/9440

But I'm sure ollama is going to get Pixtral support very soon, seeing that llama3.2 vision support is being rolled in 0.4

<!-- gh-comment-id:2432529445 --> @pbasov commented on GitHub (Oct 23, 2024): @oderwat I believe so, yes, since it's enabled by the MLX engine and llama.cpp still doesn't support it. https://github.com/ggerganov/llama.cpp/issues/9440 But I'm sure ollama is going to get Pixtral support very soon, seeing that llama3.2 vision support is being rolled in 0.4
GiteaMirror commented 2026-04-22 09:25:15 -05:00
Author
Owner
Copy Link

@Rakhsan commented on GitHub (Oct 27, 2024):

guys I ma save you some time. I learned that pixtral will COME by end of december. top secret. now stop wasting your time checking every few days

<!-- gh-comment-id:2439884288 --> @Rakhsan commented on GitHub (Oct 27, 2024): guys I ma save you some time. I learned that pixtral will COME by end of december. top secret. now stop wasting your time checking every few days
GiteaMirror commented 2026-04-22 09:25:16 -05:00
Author
Owner
Copy Link

@nonetrix commented on GitHub (Oct 30, 2024):

@oderwat I believe so, yes, since it's enabled by the MLX engine and llama.cpp still doesn't support it. ggerganov/llama.cpp#9440

But I'm sure ollama is going to get Pixtral support very soon, seeing that llama3.2 vision support is being rolled in 0.4

Seems they are switching to their own inference engine as well, or at least for vision models. Honestly just wished they made some kind of patch set for llama cpp and maintained it, probably more productive imo if llama cpp doesn't want to add them currently

<!-- gh-comment-id:2448058382 --> @nonetrix commented on GitHub (Oct 30, 2024): > @oderwat I believe so, yes, since it's enabled by the MLX engine and llama.cpp still doesn't support it. [ggerganov/llama.cpp#9440](https://github.com/ggerganov/llama.cpp/issues/9440) > > But I'm sure ollama is going to get Pixtral support very soon, seeing that llama3.2 vision support is being rolled in 0.4 Seems they are switching to their own inference engine as well, or at least for vision models. Honestly just wished they made some kind of patch set for llama cpp and maintained it, probably more productive imo if llama cpp doesn't want to add them currently
GiteaMirror commented 2026-04-22 09:25:16 -05:00
Author
Owner
Copy Link

@Leeaandrob commented on GitHub (Nov 21, 2024):

👍

<!-- gh-comment-id:2491132643 --> @Leeaandrob commented on GitHub (Nov 21, 2024): 👍
GiteaMirror commented 2026-04-22 09:25:17 -05:00
Author
Owner
Copy Link

@Tianoklein commented on GitHub (Dec 4, 2024):

PLEASE!
Allow us to compare 2 or more images via ollama API!

<!-- gh-comment-id:2518270215 --> @Tianoklein commented on GitHub (Dec 4, 2024): PLEASE! Allow us to compare 2 or more images via ollama API!
GiteaMirror commented 2026-04-22 09:25:17 -05:00
Author
Owner
Copy Link

@nonetrix commented on GitHub (Dec 5, 2024):

PLEASE! Allow us to compare 2 or more images via ollama API!

I'm pretty sure this is a thing that the LLM itself has to support correct me if I'm wrong

<!-- gh-comment-id:2520148911 --> @nonetrix commented on GitHub (Dec 5, 2024): > PLEASE! Allow us to compare 2 or more images via ollama API! I'm pretty sure this is a thing that the LLM itself has to support correct me if I'm wrong
GiteaMirror commented 2026-04-22 09:25:18 -05:00
Author
Owner
Copy Link

@Leeaandrob commented on GitHub (Dec 5, 2024):

Pixtral is avaialable for now?

Best,

+55 21 98244-8275

Em qui., 5 de dez. de 2024 às 09:12, fuggy @.***>
escreveu:

PLEASE! Allow us to compare 2 or more images via ollama API!

I'm pretty sure this is a thing that the LLM itself has to support correct
me if I'm wrong

—
Reply to this email directly, view it on GitHub
https://github.com/ollama/ollama/issues/6748#issuecomment-2520148911,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/ABL5ZOG2G7ESRYQ7JUQVJMT2EA7JRAVCNFSM6AAAAABOANEDJGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKMRQGE2DQOJRGE
.
You are receiving this because you commented.Message ID:
@.***>

<!-- gh-comment-id:2520377565 --> @Leeaandrob commented on GitHub (Dec 5, 2024): Pixtral is avaialable for now? Best, +55 21 98244-8275 Em qui., 5 de dez. de 2024 às 09:12, fuggy ***@***.***> escreveu: > PLEASE! Allow us to compare 2 or more images via ollama API! > > I'm pretty sure this is a thing that the LLM itself has to support correct > me if I'm wrong > > — > Reply to this email directly, view it on GitHub > <https://github.com/ollama/ollama/issues/6748#issuecomment-2520148911>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ABL5ZOG2G7ESRYQ7JUQVJMT2EA7JRAVCNFSM6AAAAABOANEDJGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKMRQGE2DQOJRGE> > . > You are receiving this because you commented.Message ID: > ***@***.***> >
GiteaMirror commented 2026-04-22 09:25:19 -05:00
Author
Owner
Copy Link

@frostworx commented on GitHub (Dec 5, 2024):

oh my... I suggest delaying pixtral support by one week for every useless comment here...

(sorry again, subscribers)

unsubscribing

<!-- gh-comment-id:2520420521 --> @frostworx commented on GitHub (Dec 5, 2024): oh my... I suggest delaying pixtral support by one week for every useless comment here... _(sorry [again](https://github.com/ollama/ollama/issues/6748#issuecomment-2379575418), subscribers)_ *unsubscribing*
GiteaMirror commented 2026-04-22 09:25:19 -05:00
Author
Owner
Copy Link

@wijjj commented on GitHub (Jan 23, 2025):

Is this still happening, or should I switch to vLLM?

<!-- gh-comment-id:2609615224 --> @wijjj commented on GitHub (Jan 23, 2025): Is this still happening, or should I switch to vLLM?
GiteaMirror commented 2026-04-22 09:25:20 -05:00
Author
Owner
Copy Link

@ddpasa commented on GitHub (Jan 24, 2025):

Is this still happening, or should I switch to vLLM?

I switched to vllm. Ollama team is not interested in developing ollama anymore. You see this everywhere from vulkan support, to avx vnni support to adding new vlms. vllm looks like its being actively supported.

<!-- gh-comment-id:2612049760 --> @ddpasa commented on GitHub (Jan 24, 2025): > Is this still happening, or should I switch to vLLM? I switched to vllm. Ollama team is not interested in developing ollama anymore. You see this everywhere from vulkan support, to avx vnni support to adding new vlms. vllm looks like its being actively supported.
GiteaMirror commented 2026-04-22 09:25:20 -05:00
Author
Owner
Copy Link

@Thf772 commented on GitHub (Jan 30, 2025):

Unfortunately, switching to vLLM isn't possible for everyone because Ollama has many features that vLLM doesn't (such as hosting many models at a time, loading models dynamically, or splitting the model between GPU and CPU). So I at least still hope that Pixtral will be supported in Ollama.

<!-- gh-comment-id:2624466582 --> @Thf772 commented on GitHub (Jan 30, 2025): Unfortunately, switching to vLLM isn't possible for everyone because Ollama has many features that vLLM doesn't (such as hosting many models at a time, loading models dynamically, or splitting the model between GPU and CPU). So I at least still hope that Pixtral will be supported in Ollama.
GiteaMirror commented 2026-04-22 09:25:21 -05:00
Author
Owner
Copy Link

@MrLafarge commented on GitHub (Feb 2, 2025):

It would be great to have the Pixtral support in Ollama at some point, its really missing in this wonderful project.

<!-- gh-comment-id:2629587306 --> @MrLafarge commented on GitHub (Feb 2, 2025): It would be great to have the Pixtral support in Ollama at some point, its really missing in this wonderful project.
GiteaMirror commented 2026-04-22 09:25:21 -05:00
Author
Owner
Copy Link

@PeterTucker commented on GitHub (Mar 17, 2025):

Bump

<!-- gh-comment-id:2730742463 --> @PeterTucker commented on GitHub (Mar 17, 2025): Bump
GiteaMirror commented 2026-04-22 09:25:21 -05:00
Author
Owner
Copy Link

@RobVor commented on GitHub (Apr 21, 2025):

Bump

<!-- gh-comment-id:2817964107 --> @RobVor commented on GitHub (Apr 21, 2025): Bump
GiteaMirror commented 2026-04-22 09:25:22 -05:00
Author
Owner
Copy Link

@ddpasa commented on GitHub (Apr 24, 2025):

llama.cpp now supports pixtral in their new multimodal cli: https://github.com/ggml-org/llama.cpp/pull/13065 If you want pixtral, I recommend using llama.cpp directly.

<!-- gh-comment-id:2826697722 --> @ddpasa commented on GitHub (Apr 24, 2025): llama.cpp now supports pixtral in their new multimodal cli: https://github.com/ggml-org/llama.cpp/pull/13065 If you want pixtral, I recommend using llama.cpp directly.
GiteaMirror commented 2026-04-22 09:25:23 -05:00
Author
Owner
Copy Link

@czadikem commented on GitHub (May 23, 2025):

Is this ever going to be included?

<!-- gh-comment-id:2904582628 --> @czadikem commented on GitHub (May 23, 2025): Is this ever going to be included?
GiteaMirror commented 2026-04-22 09:25:23 -05:00
Author
Owner
Copy Link

@ddpasa commented on GitHub (May 23, 2025):

Is this ever going to be included?

just use llama.cpp server: https://github.com/ollama/ollama/issues/6748#issuecomment-2826697722 Ollama is slow.

<!-- gh-comment-id:2905118273 --> @ddpasa commented on GitHub (May 23, 2025): > Is this ever going to be included? just use llama.cpp server: https://github.com/ollama/ollama/issues/6748#issuecomment-2826697722 Ollama is slow.
GiteaMirror commented 2026-04-22 09:25:23 -05:00
Author
Owner
Copy Link

@robert-mcdermott commented on GitHub (May 23, 2025):

Is this ever going to be included?

just use llama.cpp server: #6748 (comment) Ollama is slow.

Ollama is slow.

It's not that slow (2x NVIDIA A6000 w NVLink), slower than vLLM at scale, but if tuned with good hardware it's fast enough for most needs:
Image

<!-- gh-comment-id:2905185900 --> @robert-mcdermott commented on GitHub (May 23, 2025): > > Is this ever going to be included? > > just use llama.cpp server: [#6748 (comment)](https://github.com/ollama/ollama/issues/6748#issuecomment-2826697722) Ollama is slow. > Ollama is slow. It's not that slow (2x NVIDIA A6000 w NVLink), slower than vLLM at scale, but if tuned with good hardware it's fast enough for most needs: ![Image](https://github.com/user-attachments/assets/47398bc5-38aa-46f8-be88-436dec9646a4)
GiteaMirror commented 2026-04-22 09:25:24 -05:00
Author
Owner
Copy Link

@wijjj commented on GitHub (Aug 18, 2025):

gpt-oss was done faster.

<!-- gh-comment-id:3197794992 --> @wijjj commented on GitHub (Aug 18, 2025): gpt-oss was done faster.
GiteaMirror commented 2026-04-22 09:25:24 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Oct 4, 2025):

$ curl -LO https://huggingface.co/ggml-org/pixtral-12b-GGUF/resolve/main/pixtral-12b-Q4_K_M.gguf
$ curl -LO https://huggingface.co/ggml-org/pixtral-12b-GGUF/resolve/main/mmproj-pixtral-12b-Q8_0.gguf
$ cat > Modelfile <<EOF
FROM pixtral-12b-Q4_K_M.gguf
FROM mmproj-pixtral-12b-Q8_0.gguf
TEMPLATE [INST] {{ if .System }}{{ .System }} {{ end }}{{ .Prompt }} [/INST]
PARAMETER stop [INST]
PARAMETER stop [/INST]
EOF
$ ollama create pixtral:12b-q4_K_M
$ ollama run pixtral:12b-q4_K_M 
>>> hello
Hello! How can I help you today? If you have any questions or need assistance with something, feel free to let 
me know. 😊

>>> describe this image: ./picture.png
Added image './picture.png'
This image features a young puppy sitting on a stone surface outdoors. The puppy is small and fluffy, with white 
fur and bright, curious eyes. It is wearing a red collar adorned with a bell. The setting appears to be a patio 
or walkway made of large, irregularly shaped stones, suggesting an outdoor environment. The background is 
slightly blurred but shows an area that looks like the entrance to a building or a covered porch. The overall 
atmosphere of the image is cute and endearing, capturing a moment of innocence and playfulness.
<!-- gh-comment-id:3368146231 --> @rick-github commented on GitHub (Oct 4, 2025): ```console $ curl -LO https://huggingface.co/ggml-org/pixtral-12b-GGUF/resolve/main/pixtral-12b-Q4_K_M.gguf $ curl -LO https://huggingface.co/ggml-org/pixtral-12b-GGUF/resolve/main/mmproj-pixtral-12b-Q8_0.gguf $ cat > Modelfile <<EOF FROM pixtral-12b-Q4_K_M.gguf FROM mmproj-pixtral-12b-Q8_0.gguf TEMPLATE [INST] {{ if .System }}{{ .System }} {{ end }}{{ .Prompt }} [/INST] PARAMETER stop [INST] PARAMETER stop [/INST] EOF $ ollama create pixtral:12b-q4_K_M $ ollama run pixtral:12b-q4_K_M >>> hello Hello! How can I help you today? If you have any questions or need assistance with something, feel free to let me know. 😊 >>> describe this image: ./picture.png Added image './picture.png' This image features a young puppy sitting on a stone surface outdoors. The puppy is small and fluffy, with white fur and bright, curious eyes. It is wearing a red collar adorned with a bell. The setting appears to be a patio or walkway made of large, irregularly shaped stones, suggesting an outdoor environment. The background is slightly blurred but shows an area that looks like the entrance to a building or a covered porch. The overall atmosphere of the image is cute and endearing, capturing a moment of innocence and playfulness. ```
GiteaMirror commented 2026-04-22 09:25:25 -05:00
Author
Owner
Copy Link

@EnlistedGhost commented on GitHub (Oct 26, 2025):

$ curl -LO https://huggingface.co/ggml-org/pixtral-12b-GGUF/resolve/main/pixtral-12b-Q4_K_M.gguf
$ curl -LO https://huggingface.co/ggml-org/pixtral-12b-GGUF/resolve/main/mmproj-pixtral-12b-Q8_0.gguf
$ cat > Modelfile <<<EOF
FROM pixtral-12b-Q4_K_M.gguf
FROM mmproj-pixtral-12b-Q8_0.gguf
TEMPLATE [INST] {{ if .System }}{{ .System }} {{ end }}{{ .Prompt }} [/INST]
PARAMETER stop [INST]
PARAMETER stop [/INST]
EOF
$ ollama create pixtral:12b-q4_K_M
$ ollama run pixtral:12b-q4_K_M

hello
Hello! How can I help you today? If you have any questions or need assistance with something, feel free to let
me know. 😊

describe this image: ./picture.png
Added image './picture.png'
This image features a young puppy sitting on a stone surface outdoors. The puppy is small and fluffy, with white
fur and bright, curious eyes. It is wearing a red collar adorned with a bell. The setting appears to be a patio
or walkway made of large, irregularly shaped stones, suggesting an outdoor environment. The background is
slightly blurred but shows an area that looks like the entrance to a building or a covered porch. The overall
atmosphere of the image is cute and endearing, capturing a moment of innocence and playfulness.

You sir... are a saint!
Epic "mic-drop" right there. Got Pixtral-12B-Q4_K_M (and many others) working perfect thanks to your supplied chat-template and ggml-org's mmproj vision-projector!

I would shake your hand if I could!

Currently I am adding a pre-configured GGUF model/file(s) that support Ollama directly to HuggingFace and Ollama-website.
This is all possible thanks to: You, Ollama, Bartowski and a few others that put in the effort to make this possible.

<!-- gh-comment-id:3448959494 --> @EnlistedGhost commented on GitHub (Oct 26, 2025): > $ curl -LO https://huggingface.co/ggml-org/pixtral-12b-GGUF/resolve/main/pixtral-12b-Q4_K_M.gguf > $ curl -LO https://huggingface.co/ggml-org/pixtral-12b-GGUF/resolve/main/mmproj-pixtral-12b-Q8_0.gguf > $ cat > Modelfile <<<EOF > FROM pixtral-12b-Q4_K_M.gguf > FROM mmproj-pixtral-12b-Q8_0.gguf > TEMPLATE [INST] {{ if .System }}{{ .System }} {{ end }}{{ .Prompt }} [/INST] > PARAMETER stop [INST] > PARAMETER stop [/INST] > EOF > $ ollama create pixtral:12b-q4_K_M > $ ollama run pixtral:12b-q4_K_M > >>> hello > Hello! How can I help you today? If you have any questions or need assistance with something, feel free to let > me know. 😊 > > >>> describe this image: ./picture.png > Added image './picture.png' > This image features a young puppy sitting on a stone surface outdoors. The puppy is small and fluffy, with white > fur and bright, curious eyes. It is wearing a red collar adorned with a bell. The setting appears to be a patio > or walkway made of large, irregularly shaped stones, suggesting an outdoor environment. The background is > slightly blurred but shows an area that looks like the entrance to a building or a covered porch. The overall > atmosphere of the image is cute and endearing, capturing a moment of innocence and playfulness. **You sir... are a saint!** *Epic "mic-drop" right there. Got Pixtral-12B-Q4_K_M (and many others) working perfect thanks to your supplied chat-template and ggml-org's mmproj vision-projector!* I would shake your hand if I could! Currently I am adding a pre-configured GGUF model/file(s) that support Ollama directly to HuggingFace and Ollama-website. This is all possible thanks to: [You](https://github.com/rick-github), Ollama, Bartowski and a few others that put in the effort to make this possible.
GiteaMirror commented 2026-04-22 09:25:26 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Oct 26, 2025):

It's been importable for a while, it just fell off the radar until somebody mentioned it in another issue.

<!-- gh-comment-id:3449009817 --> @rick-github commented on GitHub (Oct 26, 2025): It's been importable for a while, it just fell off the radar until somebody mentioned it in another issue.
GiteaMirror commented 2026-04-22 09:25:27 -05:00
Author
Owner
Copy Link

@EnlistedGhost commented on GitHub (Oct 27, 2025):

It's been importable for a while, it just fell off the radar until somebody mentioned it in another issue.

True, however, it's a big deal to llama.cpp/Ollama dominant users as Pixtral was a total "no-go" for the longest time. (Or what felt like it, lol).

Again, thank you! A fantastic model is now yet another option for Ollama. As an Ollama dominant user myself: it meant a lot when I saw your comment a few weeks ago.

Cheers,

  • Jon

PS: Enjoying Ollama on OpenSUSE Slowroll since March 2025 :)

<!-- gh-comment-id:3453167608 --> @EnlistedGhost commented on GitHub (Oct 27, 2025): > It's been importable for a while, it just fell off the radar until somebody mentioned it in another issue. True, however, it's a big deal to llama.cpp/Ollama dominant users as Pixtral was a total "no-go" for the longest time. (Or what felt like it, lol). Again, thank you! A fantastic model is now yet another option for Ollama. As an Ollama dominant user myself: it meant a lot when I saw your comment a few weeks ago. Cheers, - Jon PS: Enjoying Ollama on OpenSUSE Slowroll since March 2025 :)
GiteaMirror commented 2026-04-22 09:25:27 -05:00
Author
Owner
Copy Link

@bluentwan commented on GitHub (Feb 10, 2026):

$ curl -LO https://huggingface.co/ggml-org/pixtral-12b-GGUF/resolve/main/pixtral-12b-Q4_K_M.gguf
$ curl -LO https://huggingface.co/ggml-org/pixtral-12b-GGUF/resolve/main/mmproj-pixtral-12b-Q8_0.gguf
$ cat > Modelfile <<EOF
FROM pixtral-12b-Q4_K_M.gguf
FROM mmproj-pixtral-12b-Q8_0.gguf
TEMPLATE [INST] {{ if .System }}{{ .System }} {{ end }}{{ .Prompt }} [/INST]
PARAMETER stop [INST]
PARAMETER stop [/INST]
EOF
$ ollama create pixtral:12b-q4_K_M
$ ollama run pixtral:12b-q4_K_M

hello
Hello! How can I help you today? If you have any questions or need assistance with something, feel free to let
me know. 😊

describe this image: ./picture.png
Added image './picture.png'
This image features a young puppy sitting on a stone surface outdoors. The puppy is small and fluffy, with white
fur and bright, curious eyes. It is wearing a red collar adorned with a bell. The setting appears to be a patio
or walkway made of large, irregularly shaped stones, suggesting an outdoor environment. The background is
slightly blurred but shows an area that looks like the entrance to a building or a covered porch. The overall
atmosphere of the image is cute and endearing, capturing a moment of innocence and playfulness.

It works! Thank you very much :)

<!-- gh-comment-id:3875107083 --> @bluentwan commented on GitHub (Feb 10, 2026): > $ curl -LO https://huggingface.co/ggml-org/pixtral-12b-GGUF/resolve/main/pixtral-12b-Q4_K_M.gguf > $ curl -LO https://huggingface.co/ggml-org/pixtral-12b-GGUF/resolve/main/mmproj-pixtral-12b-Q8_0.gguf > $ cat > Modelfile <<EOF > FROM pixtral-12b-Q4_K_M.gguf > FROM mmproj-pixtral-12b-Q8_0.gguf > TEMPLATE [INST] {{ if .System }}{{ .System }} {{ end }}{{ .Prompt }} [/INST] > PARAMETER stop [INST] > PARAMETER stop [/INST] > EOF > $ ollama create pixtral:12b-q4_K_M > $ ollama run pixtral:12b-q4_K_M > >>> hello > Hello! How can I help you today? If you have any questions or need assistance with something, feel free to let > me know. 😊 > > >>> describe this image: ./picture.png > Added image './picture.png' > This image features a young puppy sitting on a stone surface outdoors. The puppy is small and fluffy, with white > fur and bright, curious eyes. It is wearing a red collar adorned with a bell. The setting appears to be a patio > or walkway made of large, irregularly shaped stones, suggesting an outdoor environment. The background is > slightly blurred but shows an area that looks like the entrance to a building or a covered porch. The overall > atmosphere of the image is cute and endearing, capturing a moment of innocence and playfulness. It works! Thank you very much :)
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 model
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#30013
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?