SmolVLM2-256M-Video-Instruct trends on Hugging Face

HuggingFaceTB · hf · 2026-11-18

HuggingFaceTB's SmolVLM2-256M-Video-Instruct is trending on Hugging Face: a 256M-parameter image-text-to-text model with video understanding, available in transformers/onnx/safetensors formats and trained on thecauldron, Docmatix and LLaVA-OneVision-Data.

Original post →

More from Models

Models channel →