Kimi Linear 48B A3B?

Just noticed this exists, 1M context MOE with 48B par seems just like what Ive been looking for - it runs pretty damn fast too compared to Qwen 3.6 35B. after some testing it seems capable of producing *not terrible* results but it always tries to go for the minimun possible output that could

中文全文翻译正在补齐，可先查看下方英文原文。

Just noticed this exists, 1M context MOE with 48B par seems just like what Ive been looking for - it runs pretty damn fast too compared to Qwen 3.6 35B. after some testing it seems capable of producing *not terrible* results but it always tries to go for the minimun possible output that could