Engines and Tools

Qwen 3.8 27B Can Interpret Model Weights Locally

Simon Willison reports that Qwen 3.8 27B running locally can interpret neural network weights, replicating capabiliti ...

Community

Debate Erupts Over GPT-6 Astra's Weight Manipulation Claims

The technical community discusses criticisms surrounding reports that GPT-6 Astra improved Qwen by reading and writin ...

New Models

Qwen3.8-27B-NVFP4-DGX-SPARK-V2 NVFP4 Quantized 27B Model: 24GB+ VRAM

Explore the unofficial release of Qwen3.8-27B NVFP4 combined with a 64k MTP recipe for accelerated speculative decodi ...

New Models

NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF Multimodal Model: 4GB+ VRAM

Explore the GGUF quantization of NVIDIA-Nemotron-3.5-Bolt-E4B by bartowski, featuring multimodal support, specificati ...

Community

Kimi K3.1 Pro Brings Open-Weights Closer to Frontier Performance

Discover Kimi K3.1 Pro by Moonshot AI, its benchmark scores, pricing, and the community discussion on open-weights ma ...

New Models

Qwen3.8-Flash-Next-E384-FP4-Local Vision-Language Model: ~96GB Memory

Explore Qwen3.8-Flash-Next-E384-FP4-Local, an unofficial NAS-optimized MoE model with reduced experts for efficient l ...

New Models

AnyLanguageModel 1.0 Released for Apple Platforms

Discover AnyLanguageModel 1.0, a Swift library for Apple platforms combining local models and cloud APIs with unified ...

Image, Video and Audio

Kromoss 1024px Image Generation Model: 32GB+ VRAM, File List

Learn about lodestones/Kromoss, an open-weight text-to-image model fine-tuned from MOSS-Image-Base to reduce blurrine ...

New Models

DeepSeek-V4.1-Flash-1017-GGUF 1M Context GGUF Model: ~560GB Memory

Unsloth has published unsloth/DeepSeek-V4.1-Flash-1017-GGUF on Hugging Face. Check out the specs, benchmarks, and det ...

Engines and Tools

llmspeed: A Simulator to Estimate Local LLM Speeds

Discover llmspeed, an open-source tool to simulate and estimate local LLM generation speeds across various hardware c ...