The biggest local LLM on your machine is useless if it can't call a single tool, no matter how many parameters it has
…Qwen3.5 27B scores 68.5% and Qwen3.5 9B hits 66.1% and then there's a big drop: Qwen 3.5 4B drops to 50.3%, and Qwen 3.5…
Search
All sources
huggingface.co
40
xda-developers.com
26
wccftech.com
8
aws.amazon.com
5
heise.de
4
golem.de
4
neowin.net
4
amd.com
3
appleinsider.com
3
techcrunch.com
2
hardwareluxx.de
2
macrumors.com
2
Tracked topic
Alibaba
Live now
Making Qwen 3.8 27B fast on Strix Halo gfx1151
Live sources:
Level1Techs Forum
People also ask
What’s New in Qwen 3.8?
For the first time, Qwen3.8 brings a Qwen-Max-class model to open release. Built on the architectural foundation of Qwen3.5, Qwen3.8 delivers substantial gains across coding, professional work, research, and long-horizon agentic tasks. Beyond answering harder questions, Qwen3.8 is designed to carry complex, multi-step tasks through to completion with greater reliability. Qwen3.8 features the following enhancements: Core Capabilities: Comprehensive improvements across coding, professional work, research, and long-horizon agentic tasks. Agent Execution: Stronger autonomous planning and better ha
Day 0 Support for Qwen 3.8 on AMD Instinct GPUsTop stories
I split my coding work between Claude, Qwen3-Coder and Gemma 4, and it costs less than paying for one subscription
…Just this week, while building a PyGame application from scratch using Qwen3-Coder running entirely on my local machine, I realized that some larger local models are now independently capable enough to…
Paper page - BioMatrix: Towards a Comprehensive Biological Foundation Model Spanning the Modality Matrix of Sequences, Structures, and Language
…Generated by Qwen/Qwen2.5-Coder-32B-Instruct We present BioMatrix, the first multimodal foundation model that natively integrates sequences, structures, and natural language for both molecules and proteins within a single…
I finally found a local LLM I want to use every day (and it's not for coding)
…I'd seen the Qwen family come up a lot, but every time I looked into it, the conversation was always about coding benchmarks and developer workflows. So I kind of just…
Paper page - LLaDA MoE v2: Scaling Mixture-of-Experts Diffusion Language Models
…With approximately 65\% as many pretraining tokens as Qwen3, LLaDA MoE v2 approaches Qwen3 on several knowledge, reasoning, and coding benchmarks. After supervised fine-tuning alone, it outperforms SDAR Chat on seven…
I replaced my ChatGPT subscription with a local AI coding tool and haven't looked back
…It's as simple as downloading Ollama and pulling the Qwen2.5-Coder 32B model with ollama pull qwen2.5-coder:32b (use 14b instead of 32b if you have less than…
I tested Qwen3 Coder Next against four other local AI coding models, and the gap was embarrassing
…Qwen3-Coder-Next (80B MoE, 3B active) Qwen3.5-122B-A10B (122B MoE, 10B active) Devstral 2 123B (123B dense model) gpt-oss-120b (117B MoE, 5.1B active) Omnicoder-9B (9B…
Paper page - Stable-Layers: Fine-Tuning Image Layer Decomposition Models with VLM-Scored Reinforcement Learning
…Generated by Qwen/Qwen2.5-Coder-32B-Instruct We present Stable-Layers, a reinforcement learning framework that eliminates the need for paired supervision by fine-tuning a pretrained layer decomposition model using…
Paper page - Program-as-Weights: A Programming Paradigm for Fuzzy Functions
…Generated by Qwen/Qwen2.5-Coder-32B-Instruct Many everyday programming tasks resist clean rule-based implementation, such as alerting on important log lines, repairing malformed JSON, or ranking search results by…
Paper page - Off-the-Shelf LLMs as Process Scorers: Training-Free Alternative to PRMs for Mathematical Reasoning
…Generated by Qwen/Qwen2.5-Coder-32B-Instruct Selecting the best response from multiple small-model samples using a stronger scorer is a simple inference-time strategy, but fails when the small…