AnalysisAI ModelsNovember 20, 2026

SenseNova-Loop 9B GGUF runs in 8GB VRAM at 22 t/s

Read original source →huggingface.co

A community GGUF quantization of SenseNova-Loop 9B (100k context flavor) runs on an RTX 4060 at about 22 tokens/sec. The tester reports it is very concise, using roughly half the tokens of qwen3.8 for reasoning and answers.

1 source

More stories today

Open the live feed