LLaVA GGUF models
77 GGUF models in the LLaVA family, quantized by 44 uploaders. Sizes run from 374M to 34.4B parameters. The longest trained context is 512K tokens.
About the LLaVA GGUF models
LLaVA is represented on ggufmodels.com by 77 GGUF models, quantized by 44 uploaders on Hugging Face. The most downloaded are ggml llava v1.5 7b, llava 1.6 and LLaVA-NeXT-Video-7B-DPO. Parameter counts run from 374M to 34.4B, and the longest trained context window is 512K tokens.
A Q4_K_M file of a LLaVA model is between 0.2 to 20.7 GB, which is the quant most people start with. Most files come from mradermacher (16), MasakiK (4) and mys (3). The quants you will meet most often are Q4_K_M (37), Q5_K_M (32), Q8_0 (31) and F16 (31), where the number is how many models offer that quant.
Frequently asked questions
Which LLaVA GGUF should I download?
The most downloaded LLaVA GGUF here is ggml llava v1.5 7b. Pick the size that fits your GPU first, then choose a quant: Q4_K_M is the usual starting point, and Q5_K_M or Q6_K give more quality if you have memory to spare. The VRAM pages below list what fits each GPU.
How much VRAM does a LLaVA model need?
Q4_K_M files of LLaVA models on this site range from 0.2 to 20.7 GB. Add the KV cache for your context length and about 0.5 GB for the runtime, which usually means 1.5 to 3 GB on top of the file at an 8K context.
Where do LLaVA GGUF files come from?
The most common quantizers here are mradermacher, MasakiK and mys. They convert the original model and publish quantized files on Hugging Face. ggufmodels.com does not host files; every download button opens Hugging Face.