MoE and quantized LLMs are reducing active parameters and VRAM footprints, making local and workstation inference more practical.
The release of Dory1.0-4B-BF16 shows how BF16 precision redefines local deployment. It is not just a small model: about 8 GB of weights shift attention to VRAM, TCO, and quantization. For teams with workstations and small servers, the signal is clear...
A 4-billion-parameter BF16 model from NVIDIA Nemotron Labs appears on Hugging Face. The size matters less than the precision: BF16 keeps it away from aggressive quantization and places it in the local deployment space on GPUs with enough VRAM. We loo...
The model card for Tencent Hy-Memory-30B-A3B sparks debate over a slippery claim: an architecture with 3 billion active parameters out of 30 billion total would beat Opus 5 on an unspecified memory test. The issue is not the benchmark but the structu...
A three-day empirical test of GYRO-M, a Q2 quantization of Qwen Flash Next, reports q3-level stability at reduced size. The project comes from a single contributor and holds up on long contexts, but shows weaknesses on Cyrillic text. The case reopens...
NVIDIA positions Nemotron 3.5 Bolt as a vision-language model for everyday agentic tasks running locally: 74B total parameters, 9B active, native 2-bit quantization, and a 24GB footprint. It supports text, images, video, and up to 256K tokens of cont...
NVIDIA's Nemotron Bolt 3.5 74B A9B arrives Oct 15 as a 74B-parameter, 9B-active MoE compressed to 24GB using 2-bit quantization with quantization-aware training. It targets speed and intelligence balance; a 4GB E4B model follows Oct 23.
LiquidAI has introduced LFM2.5-24B-A2B, a reasoning model with 24 billion total parameters and 2.45 billion active. Built for private on-device agents, it claims to compete with Qwen3.6-35B-A3B and run up to 1.7 times faster than comparable MoEs, wit...
OpenBMB has released MiniCPM-V 4.7, a 35B-A3B multimodal model and the first MoE in the series to run smoothly on-device. Combining Lookahead Expert Scheduling, layer-wise quantization, and speculative sampling, it claims 13+ tokens/s on iPhone 17 Pr...
Precision Numerics debuts with Boole-1.7B, a 1-bit Qwen3-1.7B model at 247 MB packed weights, 13.9 times smaller than BF16. It retains 85.9% of the BF16 average score across ten tasks, outperforms 1-bit Bonsai-1.7B, and nearly matches the 1.58-bit ve...
A community ranking evaluates LLMs that can run on consumer GPUs with 10-16GB of VRAM and 26-32GB of RAM. The criteria include minimum 10 tokens/s, narrative coherence, handling of typos and nuance, and frequency of slop patterns. The results highlig...
