# “Use this model”-\>Ollama: can't pull model with Q4
**URL:**
**Author:**  [@DmitryZvonarev](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2FDmitryZvonarev)%5C
**Post date:** [May 15, 2025, 8:21am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fuse-this-model-ollama-cant-pull-model-with-q4%2F155329%2F1 "2025-05-15T08:21:49Z")
Can’t pull Q4 model from [IlyaGusev/saiga\_llama3\_8b\_gguf · Hugging Face](https://huggingface.co/IlyaGusev/saiga_llama3_8b_gguf) via “Use this model”-\>Ollama button.
All Q2, Q4, Q8, f16 exist in repository, I can download files, but “Use this model”-\>Ollama only works with Q2\_K, Q8\_0, F16. It worked month ago.
By default the model field is empty, but if I copy I will get: “ollama run [hf.co/IlyaGusev/saiga\_llama3\_8b\_gguf:](http://hf.co/IlyaGusev/saiga_llama3_8b_gguf:) **Q4\_K\_M** ”
If I run ollama pull hfco/IlyaGusev/saiga\_llama3\_8b\_gguf: **Q4\_K\_M** output is “The specified tag is not available in the repository. Please use another tag or latest”
If I change tag and run ollama pull hfco/IlyaGusev/saiga\_llama3\_8b\_gguf: **Q4\_K** output is “The specified tag is not a valid quantization scheme. Please use another tag or latest”
Is this HF issue?
---
**Author:**  [@John6666](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2FJohn6666)%5C
**Post date:** [May 15, 2025, 8:52am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fuse-this-model-ollama-cant-pull-model-with-q4%2F155329%2F2 "2025-05-15T08:52:21Z")
I’m not sure if this is an HF issue or an Ollama issue…
Either way, it’s an issue.
> **[History for packages/ollama-utils/src - huggingface/huggingface.js](https://github.com/huggingface/huggingface.js/commits/main/packages/ollama-utils/src)**
>
> Utilities to use the Hugging Face Hub API. Contribute to huggingface/huggingface.js development by creating an account on GitHub.
>
>
> \## Bug description
>
> The configuration here: https://github.com/huggingface/cha…t-ui/blob/main/docs/source/configuration/models/providers/ollama.md says that the model name should be "mistral". It should be "mistral:latest" instead. Furthermore, the auto-pulling of ollama models originally added here: https://github.com/huggingface/chat-ui/pull/1227 does not work with this model name inconsistency.
>
> \## Steps to reproduce
>
> Vanilla chat-ui setup with ollama installed. Use config from above. Download mistral "ollama pull mistral". Chat-ui unable to use ollama/mistral
>
> \## Context
> \### Logs
>
> Here's a snippet from ollama's tags:
> \`\`\`
> $\> curl http://localhost:11434/api/tags
> {"models":\[{"name":"mistral:latest","model":"mistral:latest","modified\_at":"2025-01-11T16:17:32.785621658-08:00","size":4113301824,"digest":"f974a74358d62a017b37c6f424fcdf2744ca02926c4f952513ddf474b2fa5091","details":{"parent\_model":"","format":"gguf","family":"llama","families":\["llama"\],"parameter\_size":"7.2B","quantization\_level":"Q4\_0"}},{"name":"tinyllama:latest","model":"tinyllama:latest","modified\_at":"2025-01-11T16:03:16.107607114-08:00","size":637700138,"digest":"2644915ede352ea7bdfaff0bfac0be74c719d5d5202acb63a6fb095b52f394a4","details":{"parent\_model":"","format":"gguf","family":"llama","families":\["llama"\],"parameter\_size":"1B","quantization\_level":"Q4\_0"}}\]}
> \`\`\`
>
> \### Specs
>
> \- \*\*OS\*\*: Mac
> \- \*\*Browser\*\*: Firefox
> \- \*\*chat-ui commit\*\*: f82af0b
>
> \## Notes
>
> If you change the config to:
>
> \`\`\`
> MONGODB\_URL=mongodb://localhost:27017/
> MODELS=\`\[
> {
> "name": "Ollama Mistral",
> "chatPromptTemplate": "\{{#each messages}}{{#ifUser}}\[INST\] {{#if @first}}{{#if @root.preprompt}}{{@root.preprompt}}\\n{{/if}}{{/if}} {{content}} \[/INST\]{{/ifUser}}{{#ifAssistant}}{{content}}\ {{/ifAssistant}}{{/each}}",
> "parameters": {
> "temperature": 0.1,
> "top\_p": 0.95,
> "repetition\_penalty": 1.2,
> "top\_k": 50,
> "truncate": 3072,
> "max\_new\_tokens": 1024,
> "stop": \["\"\]
> },
> "endpoints": \[
> {
> "type": "ollama",
> "url" : "http://127.0.0.1:11434",
> "ollamaName" : "mistral:latest"
> }
> \]
> }
> \]\`
> \`\`\`
> Notice the ollamaName: "mistral:latest" it works.
---
**Author:**  [@mahmutc](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2Fmahmutc)%5C
**Post date:** [May 15, 2025, 9:56am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fuse-this-model-ollama-cant-pull-model-with-q4%2F155329%2F3 "2025-05-15T09:56:12Z")
I’m not sure if this is a bug, but I noticed an inconsistency in the file type labels for these GGUF models:
1. **Model (Q2\_K):**
- Link: [saiga\_llama3\_8b\_gguf - model-q2\_K.gguf](https://huggingface.co/IlyaGusev/saiga_llama3_8b_gguf/tree/main?show_file_info=model-q2_K.gguf)
- Label: `general.file_type` → **Q2\_K**
1. **Model (Q4\_K):**
- Link: [saiga\_llama3\_8b\_gguf - model-q4\_K.gguf](https://huggingface.co/IlyaGusev/saiga_llama3_8b_gguf/tree/main?show_file_info=model-q4_K.gguf)
- Label: `general.file_type` → **Q4\_K\_M**
Is this a labeling inconsistency, or could it indicate an issue with the GGUF file format?
The **Q2\_K** file is labeled simply as `Q2_K`, while the **Q4\_K** file shows `Q4_K_M`.
Shouldn’t both follow the same naming convention (e.g., `Qx_K`)?
Could this affect compatibility or is it just a metadata oversight?
---
**Author:**  [@John6666](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2FJohn6666)%5C
**Post date:** [May 15, 2025, 10:25am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fuse-this-model-ollama-cant-pull-model-with-q4%2F155329%2F4 "2025-05-15T10:25:08Z")
There are no specific rules for file names…
Either way seems fine.
If it was working fine before (a month ago), then either Ollama or HF was pulling files according to the file name rules, or vice versa.
> **[Difference in different quantization methods · ggml-org llama.cpp ·...](https://github.com/ggml-org/llama.cpp/discussions/2094)**
>
> Hello, I'm wondering what quantization method or what you want to call it has the best output quality. Should you use q8\_0, q4\_0 or anything in between? I'm asking this question because the q8\_0 ve...
---
**Author:**  [@DmitryZvonarev](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2FDmitryZvonarev)%5C
**Post date:** [May 15, 2025, 10:28am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fuse-this-model-ollama-cant-pull-model-with-q4%2F155329%2F5 "2025-05-15T10:28:55Z")
I don’t know, but I have already successfully used q4 model.
As John6666 said, it seems like HF “Use this model” integration became more “strict” and began to check the name in the model metadata with the name of file\quantization.