AnalysisAI ModelsNovember 20, 2026

SenseNova-Loop 9B runs in 8GB VRAM via GGUF quant

Read original source →huggingface.co

A community GGUF build of SenseNova-Loop-9B fits in 8GB VRAM; one user reports ~22 tokens/sec on an RTX 4060 using the 100k context variant. They say its reasoning and answers are concise, roughly half the tokens of qwen3.8.

1 source

More stories today

Open the live feed