Prince Canuma
@Prince_Canuma
Creator of (@Nativ_AI, mlx-audio & mlx-vlm) โข working on something new
Ex-@arcee_ai โข @neptune_ai โข
๊ฐ์
July 2012
1.3K ํ๋ก์ ์ค 22.8K ํฌ
Congrats to @liquidai on LFM2.5-2.6B! Excited to have partnered with them for Day 0 support in Nativ ๐
Built for agentic + coding workflows โ and itโs fast.
On an M5 Max (48GB) with Nativ v0.2.2 โ full bf16, no quantization:
โก 11,231 tok/s prefill
โก ~84 tok/s decode
๐ง Full 128K context in just 8.5GB
๐ 476 tok/s aggregate decode at batch 16
Local inference doesn't get much better than this.
Get started today๐๐ฝ
๋ ๋ณด๊ธฐ
