Model reference · open weights

vip-llava

vip-llava is an open-weight language model from llava-hf. vip-llava-7b-hf (FP16) weighs 14.2 GB; the smallest configuration that runs it is RTX 4090 24 GB.

For self-hosting, compare this build's requirements below with the GPU you plan to rent.

• Vip-LLaVa is a 7.1B parameter open-source multimodal model developed by llava-hf for image-text-to-text tasks.
• It is an auto-regressive transformer-based chatbot that enhances the LLaVA training protocol by using visual cues such as bounding boxes and arrows.
• The model supports a context length of 4096 tokens, operates in English, and is released under the LLAMA 2 Community License.

Summary of the llava-hf/vip-llava-7b-hf model card, 2026-10-01

Hugging Face ↗Paper ↗