A Proxmox VE 9 host with two Radeon AI PRO R9700 32 GB GPUs achieved working peer-to-peer (P2P) transfers at ~13 GiB/s and ran Qwen3.8 27B UD-Q8_K_XL at ~31 t/s (non-MTP) and ~49-50 t/s (MTP n=3) usin…
A developer successfully used Unsloth to perform continued pretraining on Qwen 3 4B, teaching the small local LLM to act as a travel advisor for a fictional city by targeting only 66 million parameter…
A developer using Qwen3.8-27B from Unsloth with dynamic 3.0 quants in opencode reported that the model attempted to fetch 16 weird URLs during a single session, including a URL with an Expires timesta…
Unsloth released Dynamic 3.0 GGUFs, a quantization package that enables running the Qwen3.8-27B vision-reasoning model on 17GB of RAM, and it has been downloaded five million times in five days on Hug…
Alibaba released the open-weight Qwen3.8-27B multimodal model on 2026-08-05 under Apache-2.0, featuring a 27B dense architecture with a vision encoder, 262,144-token native context, and hybrid thinkin…
Unsloth released Dynamic 3.0, an update to its post-training quantization method for Qwen3.8-27B, claiming more than 10% better top-1 accuracy at the same disk size than competing GGUF files, based on…
Unsloth released Dynamic v3.0 GGUFs for Qwen3.8-27B, claiming more than 10% better top-1% accuracy at the same size compared to every other provider. The new quants, which work with llama.cpp and Unsl…
A developer detailed a setup for running Qwen3.8 27B, an agentic coding model, locally on a 24 GB VRAM machine using llama.cpp and the pi coding agent. The configuration achieves about 8 tokens per se…
Unsloth, the open-source fine-tuning library, released Unsloth Desktop, a free beta app for macOS, Windows, and Linux that runs and trains LLMs, diffusion models, and audio models locally, directly co…
A developer reconstructed the source code for the 1994 Windows game Rats! using the local LLM Qwen3.8 27B BF16 on a 2024 MacBook Pro with an Apple M4 Max and 128 GB of memory, achieving an average rec…
Meta Superintelligence Labs released Muse Glimmer on August 10, a 30B-parameter open-weight model under Apache 2.0, designed for local agentic workflows and tool calling. The model, available on Huggi…
Alibaba released Qwen3.8-27B, a 27.78-billion-parameter open-weights multimodal model under Apache 2.0, with a 262,144-token native context window and configurable reasoning. Vendor-reported benchmark…
Unsloth released dynamic GGUF quantizations for Qwen3.8, enabling the 27B model to run locally on 17-19GB VRAM setups and the 2.4T parameter model to run in 397GB via 1-bit quantization. The Qwen3.8 f…
DeepSeek, Alibaba, and xAI all shipped new frontier models on August 13, 2026: DeepSeek V4 Pro 0813 (1.6T MoE, 49B active, 1M context, priced at $0.435/$0.87 per million tokens), Qwen3.8-2.4T-A95B (op…
Unsloth Desktop, a new local AI training tool from Unsloth, enables users to train models locally without a cloud backend, supporting NVIDIA, AMD, Intel, and Mac hardware. It claims a 70% reduction in…
Unsloth released Unsloth Desktop (Beta), a free, open-source app for running and training AI models locally on macOS, Windows, and Linux, supporting LLMs, diffusion image/video, MLX, GGUF, and audio m…
NVIDIA introduced Nemotron 3.5 Lightning, a customizable open 30B mixture-of-experts model for always-on agents, delivering up to 4x faster token generation and 30% faster time to completion compared …
NVIDIA's open-weights Nemotron 3.5 Lightning mixture-of-experts model, with 30 billion total parameters but only 3 billion active, is designed for agentic grunt work and can be run locally on consumer…
Qubitz, a local-first AI agent for GGUF models on llama.cpp, aims to make 7B–35B MCP/tool-capable LLMs more predictable and useful through a specialized harness and Agent Behavioral Contracts. It oper…
The Watershed moment in AI is not a single event but four distinct watersheds at different scales and price points, according to a new analysis. The first, the Frontier Watershed, occurred in November…