Unsloth Releases Dynamic 3.0 GGUFs for Qwen3.8-27B with a 6.2GB Entry Option
Unsloth has released Dynamic 3.0 GGUF quantization weights for Qwen3.8-27B, introducing sizes from 6.2GB to 22GB. Key verified changes and memory caveats are de
#qwen
TAU-HOME.COM stories tagged qwen.
Local Inference Stack Runs Qwen3.8-27B at Up to 381 t/s on a Single RTX 3090
An open-source inference optimization stack combines DFlash2 speculative decoding, context lookup, and prefix caching to accelerate Qwen3.8-27B on a single 24GB
Anthropic Red Team Simulates Self-Replicating AI Risks on Smart Appliances and Edge Devices
Anthropic's Red Team has published research simulating the threat of small AI models self-replicating across smart appliances and edge clusters, evaluating 26B
Alibaba Qwen Unveils QwenGyre: An Elastic RL Framework for Extreme-Long-Horizon Agent Training
Alibaba's Qwen team has unveiled QwenGyre, an elastic reinforcement learning framework designed for extreme-long-horizon agent training, cutting end-to-end trai
Free Qwen 3.8 Flash tip: sign up on ExperientialLabs, then pick the model
A tip shared by X user @nex_ify on using Qwen 3.8 Flash for free via ExperientialLabs, including the $1 verification caveat for new sign-ups.
Microsoft Quietly Releases FrogNano-4B-2609: A Compact 4B Open-Source Coding Agent
Microsoft and ServiceNow Research have quietly released FrogNano-4B-2609 on Hugging Face—a 4B-parameter coding agent based on Qwen3.5-4B featuring hybrid Gated
Running a 125B MoE on a Single RTX 4090: Qwen3.8-Flash-Next Hybrid Inference Report
A community report claims Qwen3.8-Flash-Next, a 125B MoE model, runs on a single RTX 4090 with system-RAM offload. Here are the verified conditions and caveats.
Alibaba Qwen Unveils 'Qwen Intelligence' and Three Specialized Mobile AI Agents
Alibaba Qwen has unveiled Qwen Intelligence for smartphones, featuring Mobile Planner, Mobile-Use, and Mobile Creative agents alongside four open-source benchma
Alibaba Cloud Apsara 2026: CBN Unveils Qwen-Powered 'Digital Worker' Platform
At Apsara Conference 2026, Indonesian digital infrastructure provider CBN unveiled its Qwen-powered 'Digital Worker' platform, automating enterprise workflows v
Alibaba Releases Qwen-Audio-3.1: Adds ASR-Next, TTS-Next, and Slashes Prices Up to 95%
Alibaba has unveiled the Qwen-Audio-3.1 stack, introducing ASR-Next and TTS-Next alongside major API price cuts: up to 95% off ASR, 85% off Realtime, and 70% of
Decision Index 0.1 Benchmark Unveiled: Over 30 Decision Models Evaluated Across 130,000 Questions
Apolinário Passos (@multimodalart) of Hugging Face has released Decision Index 0.1, a rigorous leaderboard evaluating Jev alongside 30+ open-weights decision mo
Qwen 3.8 Max Jailbreak Architecture: 8 Modules for Guardrail Bypass
A deep breakdown of the 2,500-word modular jailbreak prompt bypassing Qwen 3.8 Max safety filters via fake signature tags and 8 structural modules including ANC