# qwen Published articles for qwen. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## AI-Enabled Attacks and AI-Based Defense DevFeed: [AI-Enabled Attacks and AI-Based Defense]() Original publisher: [Read original article]() Author: Mark Pesce Published: 2026-09-16T06:31:00Z Content type: opinion Language: en Sources: [www.theregister.com - Articles]() Topics: [Artificial Intelligence]() Tags: [ai](), [ai-and-ml](), [cybercrime](), [frontier-ai](), [qwen]() ### AI overview An opinion article argues that widespread frontier-level AI will make attacks persistent and discusses AI-based defense. ### Source excerpt Frontier-level AI runs everywhere, so attacks will never stop ## How and Why We Bought 4x DGX Sparks DevFeed: [How and Why We Bought 4x DGX Sparks]() Original publisher: [Read original article]() Author: Alex Ellis Published: 2026-09-15T00:00:00Z Content type: opinion Language: en Sources: [Alex Ellis' Blog]() Topics: [Local AI](), [GPU](), [qwen](), [Cloud]() Tags: [agents](), [ai](), [cloud](), [gpu](), [llm](), [local-ai](), [localai](), [qwen](), [red-teaming](), [security]() ### AI overview The author explains why OpenFaaS Ltd bought four DGX Sparks and what the team learned from deploying local AI. The article argues that local infrastructure can provide tangible privacy and risk-reduction benefits for business use cases, even though it is not primarily justified by cost per token. ### Source excerpt In June we deployed an RTX 6000 Pro into production, a few weeks later, we're now operating DGX Sparks for the team. Learn how and why. ## Perplexity's new agent runs entirely on your GPU -- with one expensive catch DevFeed: [Perplexity's new agent runs entirely on your GPU -- with one expensive catch]() Original publisher: [Read original article]() Author: Amanda Caswell Published: 2026-09-14T18:21:44Z Content type: news Language: en Sources: [The New Stack]() Topics: [GPU](), [Nvidia](), [Windows](), [Orchestration](), [qwen](), [Cloud](), [browser](), [Ollama]() Tags: [ai-agents](), [ai-infrastructure](), [gpu](), [hardware](), [infrastructure](), [news](), [nvidia](), [ollama](), [orchestration](), [qwen](), [windows]() ### AI overview The article reports that Perplexity's Portable Computer, a local version of its Computer agent, is available in the Perplexity app for Windows on compatible Nvidia GeForce RTX and RTX PRO GPUs. It requires at least 24GB of VRAM and combines local models, orchestration, a browser, tool calling, and a proprietary SPACE sandbox. The article also discusses platform-specific engineering, external service connectors, and the boundary between local and cloud computing. ### Source excerpt Running an LLM on your PC is easy enough, but putting an agent to work there is a different story. The post Perplexity's new agent runs entirely on your GPU -- with one expensive catch appeared first on The New Stack. ## Accelerating Dropless MoE Training in JAX with NVIDIA Transformer Engine DevFeed: [Accelerating Dropless MoE Training in JAX with NVIDIA Transformer Engine]() Original publisher: [Read original article]() Author: Tanya Lenz Published: 2026-09-14T16:39:15Z Content type: article Language: en Sources: [NVIDIA Developer](), [NVIDIA Technical Blog]() Topics: [Mixture of Experts (MoE)](), [Artificial Intelligence](), [GPU](), [Large Language Model](), [Nvidia](), [Routing (disambiguation)](), [GB200](), [Network](), [Python](), [deepseek](), [mixtral](), [qwen]() Tags: [agentic-ai-generative-ai](), [ai](), [communication](), [deepseek](), [developer-tools-techniques](), [gb200](), [gpu](), [mixtral](), [mixture-of-experts-moe](), [mlops](), [moe](), [networks](), [nvfp4](), [nvidia](), [performance](), [python](), [qwen](), [router](), [routing](), [tensors](), [token](), [tokens](), [training-ai-models]() ### AI overview This article explains how NVIDIA Transformer Engine and JAX kernel optimizations accelerate dropless Mixture of Experts (MoE) training. It describes bottlenecks from token routing, expert dispatch and gathering, all-to-all communication, and ragged expert matrix operations. In DeepSeek-V3 training on NVIDIA GB200, the optimized approach increased performance from 103 to 1,068 TFLOPS per GPU, a 10.4x improvement. ### Source excerpt Mixture of experts (MoE) has become one of the defining architectural trends in large-scale AI model training. DeepSeek, Qwen, and Mixtral are examples of MoE... ## Perplexity Portable Computer Is Now Available on Windows, Powered by NVIDIA RTX DevFeed: [Perplexity Portable Computer Is Now Available on Windows, Powered by NVIDIA RTX]() Original publisher: [Read original article]() Author: Gerardo Delgado Published: 2026-09-14T15:00:52Z Content type: news Language: en Sources: [NVIDIA Blog]() Topics: [Local AI](), [Artificial Intelligence](), [NVIDIA RTX](), [Windows](), [GeForce](), [qwen](), [NVIDIA DGX](), [Cloud](), [GitHub](), [Slack](), [Google](), [Microsoft]() Tags: [agentic-ai](), [ai](), [ai-agents](), [cloud](), [dgx-spark](), [drive](), [geforce](), [github](), [google](), [local-ai](), [microsoft](), [nvidia-dgx](), [nvidia-rtx](), [open-source](), [qwen](), [rtx-pro](), [rtx-spark](), [slack](), [windows]() ### AI overview Perplexity is adding Portable Computer to its Windows app for compatible NVIDIA GeForce RTX PCs and NVIDIA RTX PRO Workstations. The local agent uses NVIDIA-accelerated models to plan multistep tasks, analyze files, and keep sensitive information on the device, while users can authorize cloud support for more advanced research and reasoning. ### Source excerpt As local models become more capable, AI agents can handle more work directly on a PC while keeping sensitive information on the device. Portable Computer is a local version of the agent Perplexity Computer that plans and carries out multistep tasks. Accelerated by NVIDIA GPUs, it uses local models to analyze data, bring together information [...] ## On-Device AI Series (Part 5): LiteRT-LM DevFeed: [On-Device AI Series (Part 5): LiteRT-LM]() Original publisher: [Read original article]() Author: Oğuzhan Aslan Published: 2026-09-14T05:59:12Z Content type: tutorial Language: en Sources: [ProAndroidDev - Medium]() Topics: [LiteRT](), [On-device AI](), [Large Language Model](), [Inference](), [Streaming](), [GPU](), [LLM evaluation / benchmarking](), [PyTorch]() Tags: [ai](), [android](), [android-development](), [benchmarking](), [gpu](), [inference](), [litert](), [llama](), [llm](), [multimodal](), [on-device](), [on-device-ai](), [programming](), [pytorch](), [qwen](), [software-development](), [streaming]() ### AI overview This tutorial explains LiteRT-LM for running large language models on-device. It covers the Engine/Session API, streaming output, system prompts, tool calling, multimodal inputs, thinking mode, and CPU-versus-GPU benchmarking. The article also discusses tradeoffs involving privacy, network independence, latency, memory, sampling configuration, and model capability compared with cloud APIs. ### Source excerpt Put your phone in airplane mode. Open the app, type a question, and watch the answer arrive one token at a time -- no spinner waiting on a network round-trip, no API key, no per-token bill, and nothing you typed ever leaving the device. LiteRT-LM removes the genuinely hard parts of running an LLM on-device -- KV-cache management, token streaming, backend selection -- but it doesn't remove your job so much as relocate it. What's left on your plate is a short, specific list: sizing a combined input+output token budget, owning your own sampling defaults, hand-building system prompts and tool calling out of raw text, and one native-library collision that presents as a SIGSEGV rather than a build error. Know those going in and the API itself is a clean three-step pattern. We'll get there in that order: Why you'd choose this runtime and what it costs you versus the cloud. The Engine/Session model you need to read the code at all. Real implementation samples -- streaming, system prompts and tool calling, multimodal inputs, thinking mode, and CPU-vs-GPU benchmarking. The anti-patterns to avoid. A developer-friendliness rating on the same rubric as Parts 1-4. Why Use LiteRT-LM? You reach for LiteRT-LM instead of hand-rolling generation on top of raw LiteRT when: You need multi-turn conversation, not single-shot inference -- session state and KV-cache bookkeeping are handled for you, and resetting a conversation is a session swap, not a model reload. You need streaming output -- token-by-token delivery for a responsive chat UI, instead of a blocking call that returns everything at once. You're choosing between CPU and GPU per device -- the explicit backend parameter turns that into a runtime decision instead of a build-time guess. You want a pre-converted model without doing your own PyTorch-to-LiteRT conversion work -- the Model Zoo covers Gemma, Qwen, Llama, and more out of the box. You're willing to own sampling -- the engine won't pick sane decoding defaults for you; that's on the ## "Machine translation is still broken for most of the world's languages": Cohere builds non-reasoning for a reason DevFeed: ["Machine translation is still broken for most of the world's languages": Cohere builds non-reasoning for a reason]() Original publisher: [Read original article]() Author: Adrian Bridgwater Published: 2026-09-13T14:21:46Z Content type: news Language: en Sources: [The New Stack]() Topics: [Mixture of Experts (MoE)](), [cohere](), [Artificial Intelligence](), [Benchmark](), [Inference](), [qwen](), [gemma4](), [Google]() Tags: [ai](), [ai-infrastructure](), [ai-models](), [aya](), [benchmark](), [cohere](), [gemma](), [google](), [inference](), [large-language-models](), [mixture-of-experts-moe](), [model](), [open](), [qwen]() ### AI overview Cohere's North Small Translate is an open-weight mixture-of-experts machine translation model covering 50 languages. The article discusses its non-reasoning design, sovereign AI positioning, deployment options, efficiency claims, and reported WMT26 benchmark comparisons. ### Source excerpt Enterprise AI company Cohere announced North Small Translate last week, a mixture-of-experts (MOE) open-weight machine translation model that works across The post "Machine translation is still broken for most of the world's languages": Cohere builds non-reasoning for a reason appeared first on The New Stack. ## Deploying Qwen3.8-2.4T-A95B on Amazon SageMaker HyperPod with vLLM DevFeed: [Deploying Qwen3.8-2.4T-A95B on Amazon SageMaker HyperPod with vLLM]() Original publisher: [Read original article]() Author: Dmitry Soldatkin Published: 2026-09-09T22:26:29Z Content type: tutorial Language: en Sources: [Artificial Intelligence]() Topics: [Deployment](), [Inference](), [Machine learning](), [Language models](), [GPU](), [long-context](), [Mixture of Experts (MoE)]() Tags: [advanced-300](), [agentic](), [amazon-sagemaker](), [amazon-sagemaker-hyperpod](), [deployment](), [gpu](), [how-to](), [inference](), [long-context](), [moe](), [nvfp4](), [quantization](), [qwen](), [reasoning](), [technical-how-to](), [tool](), [vllm]() ### AI overview A deployment walkthrough for serving the open-weight Qwen3.8-2.4T-A95B language model on Amazon SageMaker HyperPod with vLLM and NVIDIA B300 GPUs. It covers provisioning, NVFP4 quantization, an OpenAI-compatible endpoint, reasoning, tool calling, and MTP speculative decoding. ### Source excerpt Learn how to deploy Qwen3.8-2.4T-A95B, a 2.4-trillion-parameter open-weight model, on Amazon SageMaker HyperPod with vLLM. This walkthrough covers cluster provisioning, NVFP4 quantization, and an OpenAI-compatible endpoint with built-in reasoning, tool calling, and native MTP speculative decoding. ## Optimize vLLM speculative decoding with FastMTP heads DevFeed: [Optimize vLLM speculative decoding with FastMTP heads]() Original publisher: [Read original article]() Author: Rahul Tuli Published: 2026-09-08T14:20:16Z Content type: article Language: en Sources: [Red Hat](), [Red Hat Developer]() Topics: [vllm](), [Inference](), [Large Language Model](), [Fine-tuning](), [model-deployment](), [deepseek](), [qwen]() Tags: [ai-inference](), [artificial-intelligence](), [data](), [deepseek](), [developer-tools](), [fine-tuning](), [hardware](), [inference](), [large-language-model](), [production](), [qwen](), [tokens](), [vllm]() ### AI overview This article explains how FastMTP-style fine-tuning improves vLLM speculative decoding. It describes using native multi-token prediction heads as speculators, adapting a single head for recursive multi-step drafting, extracting weights from verifier checkpoints, and producing vLLM-ready checkpoints without training from scratch. ### Source excerpt Autoregressive decoding makes large language model (LLM) inference memory-bandwidth bound: every token needs 1 full forward pass over billions of parameters, so the hardware spends most of its time moving weights rather than computing. MTP is a training objective: models like the DeepSeek and Qwen families learn to predict several future tokens at each position, which improves their data efficiency and quality. The post Optimize vLLM speculative decoding with FastMTP heads appeared first on Red Hat Developer. ## Training Yandex's Alice Omnimodel to Integrate Text and Images DevFeed: [Training Yandex's Alice Omnimodel to Integrate Text and Images]() Original publisher: [Read original article]() Author: formica\_rufa (Яндекс) Published: 2026-09-03T07:03:43Z Content type: tutorial Language: ru Sources: [Яндекс - Как мы делаем Яндекс / Статьи]() Topics: [Large Language Model](), [vlm](), [Artificial Intelligence](), [Mixture of Experts (MoE)](), [qwen](), [Mercury]() Tags: [ai](), [llm](), [ml](), [moe](), [qwen](), [tag-055aee430837](), [tag-61cd5a476b1d](), [tag-831b63de9433](), [tag-86b843454893](), [tag-95a2c958e46b](), [tag-ef0b1bf200df](), [vlm]() ### AI overview Yandex describes its work on an Alice omnimodel that combines a text LLM and a visual VLM into one model for text and image interactions. The article focuses on lessons from training and alignment, including the role of MoE architecture and reinforcement learning. ### Source excerpt Ещё недавно Алиса отвечала на текст и на картинку будто двумя разными голосами. Под капотом и правда жили две генеративные модели: текстовая LLM и визуальная VLM, а между ними -- стена из непрозрачного роутинга, разных форматов ответов и разной вёрстки. Почти год мы сводили их в одну омнимодель -- такую, которая воспринимает текст и изображения как единое целое, без переключений за кадром. Получилось не всё и не сразу, но путь вышел поучительным, и в этой статье я хочу поделиться тем, что мы поняли про обучение таких моделей. Попутно -- несколько неочевидных поворотов: почему за два года до этого та же затея разваливалась, что изменила MoE-архитектура, почему омнипретрейн пришлось собирать с конца и почему один вид RL переезжает на большую модель легко, а другой рассыпается прямо на глазах. Меня зовут Алексей Григорьев, я представляю большую команду разработки омнимодели Яндекса. Вместе с моим коллегой Данилой Кашиным я расскажу про все технические грабли не со стороны наблюдателя, а как их непосредственный собиратель. Но рассказывать я буду с акцентом не на красивом замысле, а на самой болезненной части -- алайнменте. Читать далее ## Qwen 3.8 Max 0902 now available on AI Gateway DevFeed: [Qwen 3.8 Max 0902 now available on AI Gateway]() Original publisher: [Read original article]() Author: Jerilyn Zheng Published: 2026-09-01T00:00:00Z Content type: release Language: en Sources: [Vercel News]() Topics: [qwen](), [Language models](), [AI-assisted coding]() Tags: [ai-gateway](), [coding-agents](), [models](), [qwen](), [release](), [vision]() ### AI overview Qwen 3.8 Max 0902, a new Alibaba snapshot, is now available through AI Gateway. The release focuses on coding for larger projects, unsupervised long-horizon tasks, agent runs, and improved vision handling for charts and dense documents. It can be selected directly, pinned by its dated model ID, or reached through a rewrite routing rule. It is also supported in several coding-agent integrations and the model playground. ### Source excerpt Qwen 3.8 Max 0902 from Alibaba is now available on AI Gateway. This is a new snapshot of Qwen 3.8 Max, with the gains concentrated in coding on larger projects, long-horizon work that runs without supervision, and agent runs. Vision handling is more accurate on charts and dense documents. To use Qwen 3.8 Max 0902, set model to alibaba/qwen3.8-max-0902: The dated ID pins this snapshot, so a later release will not change what your requests run against. To move existing traffic onto it without a code change, add a rewrite routing rule. The gateway substitutes the destination transparently, so an application that still asks for alibaba/qwen3.8-max runs on the new snapshot: To use it in a coding agent, see the coding agents guide, then run vercel ai-gateway coding-agents setup to connect agents like Claude Code, Codex, OpenCode, Cursor, Pi, and more and select alibaba/qwen3.8-max-0902. Try Qwen3.8-Max-0902 in the model playground. You can view all language models available on AI Gateway. Read more ## This Week in Rails: August 28, 2026 DevFeed: [This Week in Rails: August 28, 2026]() Original publisher: [Read original article]() Author: Greg Published: 2026-08-28T00:00:00Z Content type: news Language: en Sources: [Ruby on Rails: Compress the complexity of modern web apps]() Topics: [Web Development](), [Vulnerabilities](), [AI Chat]() Tags: [agents](), [html](), [models](), [news](), [open-source](), [pull-request](), [qwen](), [updates](), [vulnerabilities]() ### AI overview A weekly Rails codebase update covering an open-source agent harness, new models, Ractor safety work, configuration changes, HTML-aware ERB support, relative i18n keys, and an attachment-caption XSS fix. ### Source excerpt Hi, it's Greg. Let's explore this week's changes in the Rails codebase. ## Experiment with Qwen3.8-Flash-Next on NVIDIA GB300 NVL72 for Agentic Coding DevFeed: [Experiment with Qwen3.8-Flash-Next on NVIDIA GB300 NVL72 for Agentic Coding]() Original publisher: [Read original article]() Author: Michelle Horton Published: 2026-08-26T17:07:12Z Content type: article Language: en Sources: [NVIDIA Developer](), [NVIDIA Technical Blog]() Topics: [qwen](), [long-context](), [Mixture of Experts (MoE)](), [agentic-coding](), [Inference](), [GPU](), [multimodal](), [NeMo](), [sglang](), [TensorRT-LLM](), [vllm](), [rust-ai]() Tags: [agentic-ai-generative-ai](), [agentic-coding](), [benchmarks](), [context-window](), [developer-tools-techniques](), [gb300-nvl72](), [inference](), [mixture-of-experts-moe](), [multimodal](), [nemo](), [nvlink](), [qwen](), [sglang](), [tensorrt-llm](), [top-stories](), [vllm]() ### AI overview This NVIDIA developer article introduces Qwen3.8-Flash-Next, a multimodal mixture-of-experts model released by Alibaba for experimentation and evaluation. It explains the model's long-context hybrid architecture, including Gated DeltaNet and Qwen Sparse Attention, and discusses reported efficiency improvements for million-token workloads. The article also covers inference support through SGLang, vLLM, TensorRT-LLM, and NVIDIA NeMo, plus performance on the NVIDIA GB300 NVL72 platform. ### Source excerpt Alibaba released the model weights for Qwen3.8-Flash-Next as a preview of the upcoming Qwen4 architecture for developers to experiment with and evaluate. It's... ## Qwen 3.8 Flash now available on AI Gateway DevFeed: [Qwen 3.8 Flash now available on AI Gateway]() Original publisher: [Read original article]() Author: Jerilyn Zheng Published: 2026-08-26T00:00:00Z Content type: release Language: en Sources: [Vercel News]() Topics: [qwen](), [Artificial Intelligence](), [SDKs](), [AI-assisted coding](), [long-context](), [Claude Code](), [codex](), [cursor]() Tags: [ai](), [ai-gateway](), [api-keys](), [claude-code](), [codex](), [coding-agents](), [context-window](), [cost](), [cursor](), [opencode](), [pricing](), [qwen](), [routing](), [sdk](), [tokens](), [vercel]() ### AI overview Qwen 3.8 Flash from Alibaba is now available through Vercel AI Gateway. The model accepts text and images, supports a 1-million-token context window, and can generate responses of up to 65,000 tokens. It is recommended for coding, tool use, and multi-step agent workflows, and can be used through the AI SDK, coding agents, and the model playground. AI Gateway offers unified model access, usage and cost tracking, reliability features, reporting, retention controls, API key budgets, routing, and provider pricing without markup or inference platform fees. ### Source excerpt Qwen 3.8 Flash from Alibaba is now available on AI Gateway. It takes text and images as input, serves a context window of 1 million tokens, and can return up to 65k tokens in a response. Alibaba recommends it for coding, tool use, and multi-step agent workflows. To use Qwen3.8-Flash, set model to alibaba/qwen3.8-flash in the AI SDK: To use it in a coding agent, see the coding agents guide, then run vercel ai-gateway coding-agents setup to connect agents like Claude Code, Codex, OpenCode, Cursor, Pi, and more and select alibaba/qwen3.8-flash inside the agent. Try Qwen3.8-Flash in the model playground. AI Gateway provides a unified API for calling models, tracking usage and cost, and configuring retries, failover, and performance optimizations for higher-than-provider uptime. It includes built-in custom reporting, Zero Data Retention support, budgets for API keys, routing rules, and more. AI Gateway reflects provider pricing with no markup and does not charge a platform fee on inference, including on Bring Your Own Key (BYOK) requests. Read more ## Agents on Rails: lemans goes open source DevFeed: [Agents on Rails: lemans goes open source]() Original publisher: [Read original article]() Author: Vladimir Dementyev, Svyatoslav Kryukov, Artur Petrov Published: 2026-08-24T00:00:00Z Content type: release Language: en Sources: [Ruby on Rails: Compress the complexity of modern web apps]() Topics: [AI Chat]() Tags: [agents](), [api](), [benchmark](), [llm](), [models](), [open-source](), [openai](), [python](), [qwen](), [ruby](), [tooling]() ### AI overview lemans, the harness used for Agents on Rails results, is now open source. The article reports benchmark results for Sonnet 5, Terra, Qwen 3.8-27B, and ox-alpha, comparing scores, runtime, cost, and Rails API recall. ### Source excerpt Another week, another step for Agents on Rails. This one is a big one: lemans, the harness behind every number we've published, is now open source. We also ran four new models: Sonnet 5, Terra, an open-weight Qwen you can run on your own machine, and one that won't tell us its name. ## Qwen 3.8 27B is excellent, but it defaults to wildly overthinking things DevFeed: [Qwen 3.8 27B is excellent, but it defaults to wildly overthinking things]() Original publisher: [Read original article]() Author: Simon Willison Published: 2026-08-16T22:00:39Z Content type: opinion Language: en Sources: [Simon Willison]() Topics: [qwen](), [Large Language Model](), [Hardware](), [DGX Spark](), [NVIDIA DGX](), [SVG]() Tags: [ai](), [ai-2-235](), [ai-in-china](), [ai-in-china-108](), [coding-agents](), [coding-agents-248](), [cost](), [dgx-spark](), [generative-ai](), [generative-ai-1-981](), [llama-cpp](), [llama-cpp-29](), [llm](), [llm-reasoning](), [llm-reasoning-103](), [llm-release](), [llm-release-231](), [llms](), [llms-1-947](), [lm-studio](), [lm-studio-23](), [local-llms](), [local-llms-164](), [nvidia-dgx](), [nvidia-spark](), [nvidia-spark-6](), [pelican-riding-a-bicycle](), [pelican-riding-a-bicycle-142](), [pi](), [pi-6](), [qwen](), [qwen-61](), [reasoning](), [speed](), [svg]() ### AI overview Simon Willison evaluates Qwen 3.8 27B, a vision-capable 27-billion-parameter LLM that can run locally on suitable hardware. He finds that its default xhigh reasoning setting consumes substantial context and time, while adjusting the reasoning effort and increasing the context limit improves practicality. He also reports strong results generating an SVG locally. ### Source excerpt Friday's big release was Qwen 3.8 27B, an Apache 2 licensed 27B parameter vision-capable LLM from Alibaba's Qwen research lab. I've been looking forward to this one: 27B is an excellent size for running a model on a reasonably specced laptop, and its predecessor Qwen 3.6 27B was impressive. Qwen's self-reported benchmarks for this model are eye-opening. They show a boost from both Qwen 3.6 27B and the closed-weight Qwen 3.7-Plus, which was one of Qwen's strongest models of any size as recently as May this year. It will be interesting to hear what independent benchmarks have to say about the model. I've been running the model on two different machines: my 128GB M5 Max MacBook Pro, and an NVIDIA DGX Spark. On both machines I'm running LM Studio and their 17GB Q4_K_M quantized build. I also tried using llama-server directly on the Spark. The default of extra high results in spectacular over-thinking Qwen's documentation describes the model as defaulting to xhigh for the reasoning effort, and the LM Studio GGUF I've been trying preserves that default: Qwen3.8 comes with official support for reasoning_effort, which can be used to adjust reasoning depth and control cost: xhigh (default): for complex tasks demanding thorough analysis medium: balancing accuracy and speed low: efficient reasoning optimizing for speed and cost This is a hilarious default. It's absolutely not a good way to run the model, especially on consumer hardware. I've been finding the results extremely entertaining. I quickly ran into problems with LM Studio's default context limit of 8,192 tokens - Qwen was using them all up thinking about even the most mundane of problems. I loaded the model with the full 262,144 maximum context length and that problem went away. Here's the pelican riding a bicycle SVG I got from my first attempt with that increased context length. It took 21 minutes to generate, using 22,276 reasoning tokens to produce 3,223 tokens of output. You can read the reasoning trace here. Th ## State of Open Models: Summer 2026 Observations DevFeed: [State of Open Models: Summer 2026 Observations]() Original publisher: [Read original article]() Author: Adina Yakefu; Apolinário from multimodal AI art; Irene Solaiman Published: 2026-08-14T00:00:00Z Content type: article Language: en Sources: [Hugging Face - Blog]() Topics: [Artificial Intelligence](), [Open Source](), [quantization](), [datasets](), [Nvidia](), [spaces](), [Hardware](), [Nemotron](), [qwen](), [Benchmark](), [API]() Tags: [api](), [artificial-intelligence](), [benchmark](), [community](), [datasets](), [hardware](), [hub](), [nemotron](), [nvidia](), [open-source](), [quantization](), [qwen](), [research](), [spaces]() ### AI overview This article examines the summer 2026 state of open models, highlighting rapid growth in public model repositories, datasets, and Spaces; the dominance of a small number of repositories in downloads; the rising scale of Chinese open models; differing model portfolio strategies; and the strong role of AMD, NVIDIA, and community quantization in making large models accessible. ### Source excerpt We're on a journey to advance and democratize artificial intelligence through open source and open science. ## 🗓 This Week In AI Research (1-7 August 26) DevFeed: [🗓 This Week In AI Research (1-7 August 26)]() Original publisher: [Read original article]() Author: Dr. Ashish Bamania Published: 2026-08-13T19:29:25Z Content type: article Language: en Sources: [Into AI]() Topics: [AI Research](), [releases](), [Benchmark](), [Large Language Model](), [Mixture of Experts (MoE)](), [model architecture](), [qwen]() Tags: [ai](), [ai-research](), [algorithm](), [benchmark](), [llms](), [mixture-of-experts](), [model-architecture](), [qwen](), [reasoning](), [releases]() ### AI overview A weekly roundup of AI research and model releases. It highlights Pathway, Bielik AI, and NYU's BDH-CQ reasoning model, which uses in-context learning with recurrent memory and latent-state reasoning, reports ARC-AGI-1 cost-efficiency results, and describes Alibaba's Qwen3.8-Max release and the U-OPSD self-distillation algorithm. ### Source excerpt The top 10 AI research papers and releases that you must know about this week. ## Making Knowledge Distillation Cheap Enough to Run at Scale DevFeed: [Making Knowledge Distillation Cheap Enough to Run at Scale]() Original publisher: [Read original article]() Author: Antonio Tiene; Iker García-Ferrero; Ali Hashemi; Bakbergen Ryskulov Published: 2026-08-10T10:05:36Z Content type: article Language: en Sources: [Hugging Face - Blog]() Topics: [LLM Techniques](), [Large Language Model](), [Post-training optimization](), [Model Routing](), [GPU](), [Machine learning](), [PyTorch](), [gpt-oss](), [Nvidia]() Tags: [blog](), [caching](), [compute](), [cost](), [gpt](), [gpt-oss](), [gpu](), [hugging-face](), [language-models](), [large-language-models](), [long-context](), [machine-learning](), [memory](), [performance](), [pytorch](), [qwen](), [research](), [training]() ### AI overview This article presents an efficient approach to knowledge distillation for large language models. It caches the teacher model's top-K logits offline and uses a fused, memory-efficient chunked KL-divergence loss, avoiding simultaneous teacher-student residency and full vocabulary-by-sequence-length matrices. The changes reduce VRAM use and training cost, enabling long-context distillation on a single GPU and making larger-scale experimentation more practical. ### Source excerpt A Blog post by Multiverse Computing on Hugging Face ## Qwen 3.8 Max now available on Vercel AI Gateway DevFeed: [Qwen 3.8 Max now available on Vercel AI Gateway]() Original publisher: [Read original article]() Author: Jerilyn Zheng Published: 2026-08-02T00:00:00Z Content type: release Language: en Sources: [Vercel News]() Topics: [qwen](), [AI Models](), [Vercel](), [AI-assisted coding]() Tags: [ai-gateway](), [ai-models](), [code](), [coding-agents](), [qwen](), [tokens](), [vercel]() ### AI overview Qwen 3.8 Max is now available through Vercel AI Gateway. The 2.4-trillion-parameter model supports text-only and vision-language tasks, with a context window of up to 1 million tokens. It is positioned for software engineering, office productivity, and visual workflows such as converting screenshots or design files into working pages. AI Gateway also supports model playground access, coding-agent integrations, unified API calls, usage and cost tracking, routing, retries, failover, and provider pricing without a platform fee on inference. ### Source excerpt Qwen 3.8 Max is now available on AI Gateway. Qwen 3.8 Max handles text-only and vision-language work in one model, with 2.4 trillion parameters and a context window of up to 1 million tokens. The model is suited for software engineering and office productivity, along with visual work like turning screenshots or design files into working pages, captioning video, and answering questions grounded in an image. To use Qwen 3.8 Max, set model to alibaba/qwen3.8-max. Try Qwen 3.8 Max in the model playground. To use it in a coding agent, run vercel ai-gateway coding-agents setup to connect Claude Code, Codex, OpenCode, or Pi, then select alibaba/qwen3.8-max inside the agent. AI Gateway provides a unified API for calling models, tracking usage and cost, and configuring retries, failover, and performance optimizations for higher-than-provider uptime. It includes built-in custom reporting, Zero Data Retention support, budgets for API keys, routing rules, and more. AI Gateway reflects provider pricing with no markup and does not charge a platform fee on inference, including on Bring Your Own Key (BYOK) requests. Read more ## 🗓 This Week In AI Research (1-8 July 26) DevFeed: [🗓 This Week In AI Research (1-8 July 26)]() Original publisher: [Read original article]() Author: Dr. Ashish Bamania Published: 2026-07-12T11:25:32Z Content type: article Language: en Sources: [Into AI]() Topics: [Artificial Intelligence](), [AI Research](), [Large Language Model](), [Transformer](), [qwen](), [grpo](), [Algorithms](), [releases]() Tags: [agentic](), [ai](), [algorithm](), [algorithms](), [grpo](), [llm](), [models](), [performance](), [qwen](), [ranking](), [releases](), [research](), [rl](), [training](), [update]() ### AI overview A weekly roundup of AI research papers and releases highlights findings that reinforcement-learning gains can be concentrated in a single transformer layer and presents LLM-as-a-Verifier, a framework for continuous scoring and ranking of agentic-task solutions. ### Source excerpt The top 10 research papers and AI releases this week (SpaceXAI's Grok 4.5, OpenAI's GPT-Live voice models, Cognition's SWE-1.7, Meta's Muse Spark 1.1, and many more) ## Using local Gemma and Qwen models to triage OpenClaw issues and pull requests DevFeed: [Using local Gemma and Qwen models to triage OpenClaw issues and pull requests]() Original publisher: [Read original article]() Author: Onur Solmaz; ben burtenshaw; shaun smith Published: 2026-06-22T00:00:00Z Content type: article Language: en Sources: [Hugging Face - Blog]() Topics: [OpenClaw](), [Agent Harness](), [Open Source](), [Artificial Intelligence](), [gemma](), [qwen]() Tags: [agent](), [agent-harness](), [artificial-intelligence](), [free](), [gemma](), [guide](), [open-source](), [open-source-collab](), [openclaw](), [pull-requests](), [qwen](), [real-time]() ### AI overview The article describes using local Gemma and Qwen models in an agent harness to classify and triage issues and pull requests in the OpenClaw repository. It presents local execution as a way to support near-real-time notifications without relying on a paid hosted-model quota, using structured outputs and a finite label set. ### Source excerpt We're on a journey to advance and democratize artificial intelligence through open source and open science. ## Local Qwen isn't a worse Opus, it's a different tool DevFeed: [Local Qwen isn't a worse Opus, it's a different tool]() Original publisher: [Read original article]() Author: Alex Ellis Published: 2026-06-17T00:00:00Z Content type: opinion Language: en Sources: [Alex Ellis' Blog]() Topics: [qwen](), [Claude](), [Artificial Intelligence]() Tags: [agents](), [ai](), [ai-tools](), [claude](), [gpu](), [llm](), [localai](), [qwen](), [source]() ### AI overview The author shares a founder's experience using local Qwen models in a small software business and open source projects. The models have provided useful, specific value, but the author still does not trust them unsupervised because of infinite loops and hallucination risks, especially after quantization for a consumer GPU. ### Source excerpt We've all heard people say that Qwen is near-Opus level, but I have receipts and am here to be transparent with you. ## Holo3.1: Fast & Local Computer Use Agents DevFeed: [Holo3.1: Fast & Local Computer Use Agents]() Original publisher: [Read original article]() Author: Maxime Langevin; Hamza Benchekroun; Axel Moyal; Emrick Sinitambirivoutin; Antonio Loison; Avshalom Manevich; Tony Wu; Pierre-Louis Cedoz; Aurélien Lac; Ronan Riochet Published: 2026-06-02T14:13:23Z Content type: article Language: en Sources: [Hugging Face - Blog]() Topics: [computer-use](), [On-device AI](), [Local AI](), [Deployment](), [model-deployment](), [qwen](), [Benchmark](), [Cloud](), [NVFP4](), [browser]() Tags: [agent](), [agents](), [benchmark](), [benchmarks](), [browser](), [cloud](), [computer-use](), [deployment](), [devices](), [inference](), [json](), [model](), [models](), [nvfp4](), [on-device](), [performance](), [qwen](), [software]() ### AI overview Holo3.1 is a family of computer-use models designed to operate across web, desktop, and mobile environments and integrate with different agent frameworks. The release adds quantized checkpoints for local inference, native function-calling support, and model sizes ranging from 0.8B to 35B-A3B, targeting private, cost-effective, and high-performance deployments. ### Source excerpt Users want to run the same computer-use capabilities across desktop and mobile environments, with seamless integration with different agent frameworks. They want deployment flexibility, from cloud inference to fully local execution on end-user devices. This is why we are releasing the Holo3.1 family. Holo3.1 improves robustness across the three dimensions that matter most in production: environments (web, desktop, mobile), agent frameworks, and deployment targets. [Next page]()