#local-inference

The latest AI and developer news about #local-inference, with the original source and a short summary.

On an RTX 3090, Qwen3.5-27B with DFlash+DDTree recorded up to 207 tok/s.

·1·2026.04.21 03:00

On M5pro 64GB, Qwen 3.5 122B ran fairly well at IQ2-m.

·2026.04.14 03:00

LazyMoE aims for ultra-low-memory execution of a 120B LLM using lazy expert loading and TurboQuant.

·1·2026.04.13 04:00

Waypoint-1.5 extends real-time world models for consumer GPUs with 720p/60FPS and 360p tiers.

·2026.04.09 09:00

Musinsa reviewed translation quality and cost using TranslateGemma 27B(Q6) instead of GPT-4o-mini.

·1·2026.03.16 07:00