# What is GGUF? Definition & Technical Mechanics

> **Source:** [SPIDITS AI AI Glossary](https://spidits.com/ai-glossary)  
> **Category:** Model Operations  
> **Canonical Citation:** [https://spidits.com/ai-glossary/gguf](https://spidits.com/ai-glossary/gguf)

## Definition
GGUF (GPT-Generated Unified Format) is a file format designed for storing models for inference with llama.cpp. It is optimized to support fast on-device loading and quantization.

## Architectural Mechanics & Deep Dive
GGUF (GPT-Generated Unified Format) is a file format designed for distributing and running quantized LLMs on consumer-grade hardware, particularly using llama.cpp. GGUF stores model weights, metadata, and vocabulary in a single file, supporting fast loading, memory mapping (mmap), and CPU/GPU split execution, making it the standard format for local open-source model deployment.

## Industry Applications & Real-World Use Cases
On-device local LLM execution, CPU-based inference, and desktop AI applications.

## Frequently Asked Questions

### Why is GGUF popular for running models locally?
It stores all metadata and weights in a single file and supports CPU offloading, allowing models to run on consumer-grade laptops.

### What format did GGUF replace?
GGML, which was deprecated due to lack of flexibility and formatting complexity.

## Related Terminology
- [quantization](https://spidits.com/ai-glossary/quantization)
- [llama](https://spidits.com/ai-glossary/llama)
- [edge-ai](https://spidits.com/ai-glossary/edge-ai)

---
*From the SPIDITS AI Knowledge Base (336+ terms indexed). Explore full technical definitions: [https://spidits.com/ai-glossary](https://spidits.com/ai-glossary)*