# GGUF download: choose and check a model file

<a id="short-answer"></a>

To download a GGUF model, choose an exact publisher and version, select a file your runner supports, and check its size and required companion files. After downloading, compare its SHA-256 with the recorded digest before loading it.

Sources checked: 2026-10-26

<a id="choose-format"></a>

## GGUF or Safetensors?

Use the format your app and the exact model architecture support. Changing a filename does not convert a model.

| Format | Where to check support | What to inspect |
| --- | --- | --- |
| GGUF | llama.cpp-based runners and supported Ollama imports | Architecture, quantization, chat template and any extra files |
| Safetensors | The model card and your inference library | Configuration, tokenizer, weight shards and supported architecture |

<a id="choose-exact-file"></a>

## Choose one exact model release

Search [GGUF files](/browse?q=GGUF), or include the full model name, such as [Qwen2.5 GGUF](/browse?q=Qwen2.5%20GGUF). Open the record and upstream model card. Quantizations are different files: compare size and your runner's supported formats rather than assuming a quantization name is a quality score. A source-only record does not promise a Hugging Bay download.

A filename like `model-00001-of-00003.gguf` is one part of a split set. Keep all three matching parts together. A vision model may also need a projector; follow its model card and runner instructions.

<a id="check-download"></a>

## Check the downloaded bytes

Save the file path, size, publisher and recorded SHA-256 from the same current file row. On macOS, compute a local digest with:

```sh
shasum -a 256 /path/to/model.gguf
```

On Linux, use `sha256sum /path/to/model.gguf`. Compare the complete digest with a trusted record. A match checks the bytes; it does not prove model quality, permission or successful execution. Then follow the [local Llama import example](/learn/run-llama-locally) or your runner's instructions.

<a id="learn-sources"></a>

## Sources

- [Hugging Face GGUF format guide](https://huggingface.co/docs/hub/gguf)
- [Import a model into Ollama](https://docs.ollama.com/import)
- [llama.cpp supported model files](https://github.com/ggml-org/llama.cpp/blob/master/docs/models.md)
- [llama.cpp GGUF split and merge implementation](https://github.com/ggml-org/llama.cpp/blob/master/tools/gguf-split/gguf-split.cpp)
- [Hugging Face model cards](https://huggingface.co/docs/hub/model-cards)

[Browse GGUF model records](/browse?q=GGUF)