Qwen

NVIDIA has published Qwen3.8-Flash-Next-FP4-Local, a 96 GiB NVFP4 quantization of Alibaba's 125B-parameter MoE model, on Hugging Face. NVIDIA reports 91.5% on GPQA Diamond versus 92.2% for BF16. It requires NVIDIA's custom vLLM containers because upstream support has not landed.

Nick Cammarata of Goodfire published an essay, "Neural Networks Are Already Clean," arguing that agentic analysis resolves many known cases of polysemanticity and superposition. A widely shared summary says 90% of neurons in Qwen-32B have a single function. The results are unreplicated, and key methodology was not available in the source material.

Qwen has published Qwen-Image-2.1-Turbo on Hugging Face, an accelerated checkpoint of Qwen-Image-2.1 that runs text-to-image generation and image editing in 8 denoising steps. It keeps the same 7B visual generation architecture and loads through a new QwenImage21Pipeline in Diffusers.

Cloudflare has released Clef and Clef-flash, two open-weight decision models that return probabilities over predefined answer options instead of generating text. The company claims median latencies of 39 ms and 209 ms, against just over 524 ms for TypeSafe AI's Jev. Both support text and images and are API-compatible with Jev.

Apple has released LensVLM-9B, a 9-billion-parameter vision-language model fine-tuned from Qwen3.5-9B-Base that processes documents as compressed images, selectively expanding only relevant pages to full resolution. The model supports 5x, 10x, and 15x compression ratios and is available under Apple's Machine Learning Research Model License.

Alibaba's Qwen team has launched Qwen3.8 Max Prime, a separate high-throughput SKU of its flagship Qwen3.8 Max model. It carries the same 1M-token context window and multimodal capabilities as the base model but at double the price.

Alibaba's Qwen team has released Qwen-Audio-3.1, a five-model lineup covering speech recognition, text-to-speech, and real-time voice interaction. Alongside the release, Alibaba cut API pricing by up to 95 percent for ASR, 85 percent for real-time models, and 70 percent for TTS.

A briefing prepared for Congress by AI researcher Nathan Lambert details how Chinese open-weight models have overtaken American counterparts since mid-2025, with a nearly 2x lead in Hugging Face downloads and a 19-22 point gap on the Artificial Analysis Intelligence Index.

Alibaba's Qwen team has open-sourced Qwen-Image-2.1, a 7B parameter unified model for text-to-image generation and image editing. The release adds native transparent (RGBA) image support and editing with up to 10 reference images.

Alibaba's Qwen team released Qwen3.8-Omni-Flash, a multimodal model for AI agents that processes audio and video with a 1 million token context window. Pricing undercuts Google's Gemini 3.8 Flash by roughly 5x on input and 8x on output, according to Qwen.

PrismML released Bonsai 2 27B, a compressed version of Alibaba's Qwen3.8 27B model that shrinks memory footprint by 9x to 10x down to 5.9GB. The startup claims 98% aggregate benchmark parity with the original, up from 95% in its first release, using a ternary weight compression technique.

Anthropic's latest threat intelligence report covers December 2025 through August 2026, documenting Claude's misuse in espionage, weapons development, and nationwide surveillance operations. The report also details how seven Chinese AI labs ran covert networks—some routing their own customers' requests through Claude—to extract training data at industrial scale.

Alibaba's Qwen team released Qwen3.8-2.4T-A95B on August 12, 2026, the open-weight version of Qwen3.8-Max and the first Qwen-Max-class model made publicly available. The 2.4 trillion-parameter mixture-of-experts model activates only 95 billion parameters per token and supports context windows up to 1 million tokens.

Alibaba has released Qwen-Drive 1.0, a driving model built on Qwen3.5-4B that handles spatial perception, route planning, and cockpit dialogue in a single system. Reinforcement learning cut the rate of off-road driving errors in simulation from 24 percent to 12 percent, though the model's stated reasoning doesn't always match its actual maneuvers.

Alibaba's Qwen team released Qwen3.8 Max (0902), a 2.4-trillion-parameter mixture-of-experts model with a 1M-token context window that accepts text, image, and video input. The snapshot is post-trained for coding, agentic workflows, and long-horizon task execution, priced at $2/$6 per 1M input/output tokens.

Perplexity's Mac app now supports Hybrid Compute, which starts tasks in the cloud and shifts sensitive steps to a local model running on-device. The feature requires Apple silicon with at least 24GB of unified memory and uses an open-sourced on-device PII classifier to mask private data before any cloud processing.

Alibaba has released Qwen3.8 Flash, a multimodal reasoning model with a 1 million token context window, aimed at coding, agentic workflows, and visual/document analysis. It's priced at $0.16 per 1M input tokens and $0.47 per 1M output tokens through Alibaba Cloud International.

Alibaba's Qwen team has released Qwen3.8-Flash-Next, an open-weight model with 125B total parameters (6B activated) that previews architectural changes planned for Qwen4, including a new sparse attention mechanism and n-gram embeddings. The model natively supports 262,144 tokens of context, extensible to 1 million.

Alibaba's Qwen research lab has released Qwen 3.8 27B, an Apache 2.0 licensed, vision-capable model with a 262,144-token context window. Independent testing found the model's default 'xhigh' reasoning setting causes it to massively overthink simple prompts, turning quick tasks into 20-minute ordeals.

Hugging Face repositories for Qwen3.8-27B, a matching FP8 quantized build, and LiquidAI's LFM2.5-VL-3B surfaced within the same news cycle. Neither Alibaba's Qwen team nor LiquidAI has published accompanying benchmarks, pricing, or technical reports as of this writing.

Alibaba's Qwen team has released FP8-quantized weights for Qwen3.8-27B, a 27-billion-parameter dense vision-language model with native 262,144-token context extensible to 1 million tokens. The model claims gains over its Qwen3.6 and Qwen3.7 predecessors on coding, agentic, and multimodal benchmarks.

Alibaba's Qwen team has released Qwen3.8-27B, a 27-billion-parameter dense vision-language model with 262,144-token native context extensible to 1 million tokens. The model shows gains over Qwen3.6-27B and Qwen3.7-Plus across coding, agentic, and multimodal benchmarks, according to Alibaba.

Hugging Face repositories for Qwen3.8-2.4T-A95B, its FP8 variant, and NVIDIA's Nemotron-3.5-Lightning-30B-A3B have surfaced, but neither company has published accompanying benchmarks, technical reports, or pricing.

Alibaba's Qwen team has released Qwen3.8-2.4T-A95B-FP8, an open-weight, FP8-quantized MoE model with 2.4 trillion total parameters and 95 billion activated per token. It natively supports 262,144 tokens of context, extensible to 1,010,000, and forms the base for the hosted Qwen3.8-Max API.

Alibaba's Qwen team has released Qwen3.8-2.4T-A95B, a 2.4-trillion-parameter mixture-of-experts model activating 95B parameters per token. The company claims it is the first open-weight release to reach Qwen-Max-class performance, with a hosted variant, Qwen3.8-Max, offering 1M-token context and vision input.

Apple has confirmed that its new Siri AI features in China will be powered by Alibaba's Qwen models rather than Google Gemini, which handles Siri AI elsewhere. A published guide indicates users will need separate Qwen accounts, suggesting the privacy protections built into Apple's Private Cloud Compute won't apply.

Alibaba's Qwen3.8 Max jumps 10 points to 56 on the Artificial Analysis Intelligence Index, putting it on par with Claude Opus 4.8. But Kimi K3 still edges it out at a lower per-task cost, and Qwen3.8 Max shows a sharp rise in hallucination rate.

Alibaba is promoting its new Qwen 3.8 model with marketing that frames AI automation as liberating rather than threatening, a departure from the fear-based messaging common among Western AI labs. The company has not disclosed technical specifications, benchmark scores, or pricing for the model.

Alibaba has released Qwen3.8-Max, a 2.4-trillion-parameter model with 95 billion active parameters per query, designed to run autonomous tasks over multiple days. The company claims it hits 93 on PaperBench and rivals Claude Opus 4.8 and GPT-5.6 Sol on internal benchmarks, with open weights arriving next week.

Alibaba has released Qwen3.7 Flash, a vision-language reasoning model with a 1 million token context window aimed at multimodal agents, visual coding, and computer-use tasks. The model is priced at $0.03 per 1M input tokens and $0.13 per 1M output tokens and is available through OpenRouter.

Alibaba's Qwen team has released Qwen-Image-3.0, an image generator that accepts prompts up to 4,500 tokens and can render legible text as small as ten pixels, complex LaTeX formulas, and twelve languages in a single pass. The model is currently invite-only via API, and unlike its predecessor, it likely won't ship with open weights.

China's Cyberspace Administration approved Apple Intelligence for launch in the country, backed by integration of Alibaba's Qwen AI model across Apple's operating systems. The deal ends a two-year delay that began when Apple Intelligence debuted in 2024.

Alibaba's Qwen team released Qwen-AgentWorld-35B-A3B, a 35 billion parameter language world model designed for agentic environment simulation. The model covers seven domains—MCP tool calling, Search, Terminal, Software Engineering, Android, Web, and OS—in a single model with a 262,144 token context window.

Alibaba's Qwen has released Qwen3.7 Plus, a multimodal model with a 1 million token context window. The model accepts text and image input with text output, priced at $0.40 per million input tokens and $1.60 per million output tokens through OpenRouter's API.

Alibaba has released Qwen3.7 Max, the flagship model in its Qwen3.7 series, featuring a 1 million token context window. The text-only model is designed for agent-centric workloads with strengths in coding, office productivity, and long-horizon autonomous execution, and includes explicit prompt caching support.

Microsoft Research has released Fara-7B, a 7-billion parameter small language model designed for computer automation tasks. The model, which took 2.5 days to train on 64 H100 GPUs, can navigate websites to complete tasks like booking restaurants and shopping, using screenshots as input with a 128K token context window.

Alibaba's Qwen team has released Qwen3.6 Flash, a multimodal language model supporting text, image, and video input with a 1 million token context window. The model is priced at $0.25 per 1M input tokens and $1.50 per 1M output tokens, with tiered pricing above 256K tokens.

Qwen has released three models in its Qwen3.6 series: a flagship Max Preview model, a 35B parameter A3B variant, and a 27B parameter base model. All three models are now accessible through OpenRouter's API platform.

Alibaba Cloud has released Qwen3.6-35B-A3B, an open-weight sparse mixture-of-experts model with 35 billion total parameters but only 3 billion active parameters per token. The model features a 262K native context window (expandable to 1M tokens), multimodal input support, and integrated reasoning mode with preserved thinking traces.

Alibaba Cloud has released Qwen3.6 Max Preview, a proprietary frontier model built on sparse mixture-of-experts architecture with approximately 1 trillion total parameters. The model supports a 262,144-token context window and features integrated thinking mode for multi-turn reasoning, priced at $1.30 per million input tokens and $7.80 per million output tokens.

Alibaba's Qwen Team has released Qwen3.6 27B, a 27-billion parameter multimodal language model with a 262,144-token context window. The model accepts text, image, and video inputs and includes a built-in thinking mode for extended reasoning, with pricing at $0.195 per million input tokens and $1.56 per million output tokens.

Alibaba announced Friday that its Qwen AI model will be integrated into vehicles from 10 Chinese automakers including BYD, Geely, Li Auto, and SAIC Volkswagen. The system runs on Nvidia's automotive chip platform and allows drivers to order food delivery, book hotels, and make payments through voice commands, even with limited network connectivity.

Alibaba Cloud released Qwen 3.6 27B, a 27-billion parameter language model, alongside an FP8 quantized version for deployment efficiency. Separately, OpenAI published a privacy filter model on Hugging Face, marking a rare public model release from the company.

Alibaba Qwen has released Qwen3.6-35B-A3B, a 35-billion parameter mixture-of-experts model with 3 billion activated parameters and a 262,144-token native context window extendable to 1,010,000 tokens. The model scores 73.4 on SWE-bench Verified and features FP8 quantization with performance metrics nearly identical to the original model.