Exceptional Release, This is The Most Powerful OSS Model for 24GB Cards
#29
by
AaronFeng753
- opened
I assumed Qwen3 32B would remain the top dense model for 24GB cards, but I was completely mistaken.
This model significantly outperforms Qwen3 32B across the board. I've tested it on coding tasks, translation, and even history trivia – it handled everything with impressive accuracy.
In many ways, it rivals certain closed-source commercial models.
Thank you for this outstanding release. I genuinely hope you continue delivering models of this caliber.
Just wanted to add my thanks too