#llm-quantization

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

Nota compresses 30B LLM to 10GB with 1.58-bit ternary quantization to run on MacBook

Reducing the 30B model size to 10GB and recovering 89.4% of benchmark scores makes it possible to run on a 24GB MacBook.

·2026.11.09 09:00

PACE: Post-hoc Four-Bit State Quantizer for Linear-Attention LLMs

PACE uses periodic rounding and channel-aligned scaling to maintain accuracy at four bits, matching 8-bit performance on models up to 35B parameters.

·1·2026.10.23 14:00

Dynamic Quantiser Released, Improving Accuracy Over Standard Quantization in VRAM-Constrained Environments

The open-source dynamic quantization tool 'Dynamic Quantiser' has been released, offering higher accuracy than standard quantization in VRAM-constrained environments for llama.cpp-compatible GGUF files.

·2026.09.22 23:00

Unsloth Dynamic 3.0 GGUFs

Unsloth has released Dynamic v3.0 quantized models for Qwen3.8-27B, demonstrating improved accuracy.

·1·2026.08.20 03:00

ExTernD: Overcoming the Accuracy Limits of Ternary Quantization

A new PTQ methodology is proposed that achieves efficient Ternary Quantization while minimizing accuracy loss through matrix decomposition.

·1·2026.07.16 22:00

Advanced Quantization Algorithm for LLMs

AutoRound has strengthened 2-4 bit LLM/VLM quantization and ecosystem integration.

·2026.05.02 00:00

Intel AutoRound Expands Quantization Support

AutoRound has expanded CPU/XPU/CUDA support and low-bit quantization formats.

·1·2026.05.01 23:00