On: Muse Glimmer: 30B-parameter model optimized for always-on local agent workflows

[segmondy]

Many people don't have $5000 for DGX Sparks. With that said, it doesn't take much to run Deepseek. I run it on a sub $1000 system 128gb 2 3060 at 6-7tk/sec and then on a $1000 system with 10 MI50 GPUs built when the price was cheap.