August 18, 2026
••
Latest News
Latest news from Subscribe to LocalLLaMA's sources
August 18, 2026
Qwen 4 expected for September
August 18, 2026
Running DeepSeek V4 Flash Q4_K_XL at ~100 tok/s prompt processing on 4× RTX 3060 12GB
August 18, 2026
If the weights never change, is it really recursive self-improvement?
August 18, 2026
Is Ling 3 tiny underrated for its size?
August 18, 2026
What sandbox are you all using for AI agents?
August 18, 2026
Agent on CPU, which to pick?
August 18, 2026
OpenCode overrides the samplers for Qwen models to the wrong values
August 18, 2026
Linux Improves VRAM Management in 7.3 Kernel 🥳
August 18, 2026
Any ~4B models with decent programming knowledge? NO VIBECODE/ONE SHOT
August 18, 2026
Hugging Face just surpassed 3 million models on the Hub
August 18, 2026
What would make an uploader-run refusal table independently reproducible?
August 18, 2026
DeepSeek V4 Flash 0731 on Strix Halo: draft model, n_max sweep, and a launch line that actually helps
August 18, 2026
AltRouter, an OpenRouter alternative
August 18, 2026
Don't ignore llama.cpp RPC with old hardware. Results of a 5070 Ti and 1080 Ti over gigabit ethernet: it's actually functional.
August 18, 2026
Qwen 3.8 27b saved me $650+ in API costs this evening
August 18, 2026
Qwen 3.8 27B is faster than expected
August 18, 2026
Qwen 3.8 27b vs Deepseek Flash
August 18, 2026
tencent/UI-Mate-27B · Hugging Face
August 18, 2026
