DFlash 2 Speeds Up Local AI by Up to 4.6 Times

Local AI just got a major speed boost, and the new tech is free. Inco AI and Z Lab say their DFlash 2 generation doubles local AI speed by up to 4.6 times.

The headline number: Qwen3.8-27B runs at 70 tokens per second on an M5 Max MacBook Pro.

What Happened

Inco AI, working with Z Lab, launched DFlash 2, the latest version of its generative technology. The team says it doubles the speed of local AI by up to 4.6 times. They demonstrated it running the Qwen3.8-27B model at 70 tokens per second on an M5 Max MacBook Pro, calling the result a legendary leap. The announcement presents DFlash 2 as a go-to tool for developers.

Why It Matters

Token speed is the bottleneck for local AI. If DFlash 2 really delivers that 4.6x gain, developers can run large models like Qwen3.8-27B at usable speeds on a laptop, without paying for cloud GPUs. That makes local development more practical for a lot of teams. The post does not include setup instructions, so the how-to is still to come.