Qwen3.8-Max release notes / what's new
Qwen has released Qwen3.8-Max, a 2.4 trillion parameter model designed for autonomous coding, professional workflows, and long-horizon tasks, with open weights arriving next week.
Qwen-Image-3.0 Release Notes
Qwen-Image-3.0 is a third-generation image generation model focused on realism and utility, featuring support for 4.5k token inputs for complex layouts, 10px small text rendering, and native support for 12 languages.
Qwen-AgentWorld release: language world model for seven domains and its impact on general agents
Qwen releases Qwen‑AgentWorld, a language world model that simulates seven agent environments and improves general agents via controllable simulation and unified next‑state prediction.
Qwen Robot Suite: Unified Foundation Models for Navigation, Manipulation, and World Modeling
Qwen introduced the Qwen‑Robot Suite—three foundation models (RobotNav, RobotManip, RobotWorld) that translate language into navigation, manipulation, and world‑prediction actions, enabling unified agentic robotics across dozens of embodiments.
Qwen-RobotWorld: Boundless Worlds for Embodied Agents
Qwen-RobotWorld is a unified world model that uses natural language as a universal action interface to enable cross-scenario physical generalization across 20+ robot embodiments.
Qwen-RobotManip: Alignment Unlocks Scale for Robotic Manipulation Foundation Models
Qwen-RobotManip is a Vision-Language-Action (VLA) foundation model that uses a unified alignment framework and a human-to-robot data synthesis pipeline to achieve state-of-the-art generalization across diverse robot embodiments and out-of-distribution tasks.
Qwen-RobotNav: A Scalable Navigation Model for Agentic Systems
Qwen-RobotNav is a unified navigation model based on Qwen3-VL that achieves state-of-the-art performance across five navigation domains by treating visual context as a controllable inference-time interface.
Qwen3.7-Plus release notes / what's new
Qwen3.7-Plus is a multimodal agent model that unifies vision and language to operate across GUI and CLI environments for complex software engineering and productivity automation.
Qwen-VLA: Unifying Vision-Language-Action Modeling for Embodied Intelligence
Qwen-VLA is a general-purpose Vision-Language-Action model that unifies robotic manipulation, vision-language navigation, and cross-embodiment control into a single generalist policy model.
Qwen3.7-Max agent model release
Qwen released Qwen3.7-Max, a new agent-focused foundation model that excels at coding, office automation, and ultra-long-horizon autonomous tasks, now available via Alibaba Cloud Model Studio.
Qwen3.5-LiveTranslate-Flash Release Notes
Qwen3.5-LiveTranslate-Flash is a simultaneous interpretation model built on Qwen3.5-Omni that provides real-time, multimodal translation across 60 languages with ultra-low latency and voice cloning.
Qwen-Scope Interpretability Toolkit Release
Qwen has released Qwen-Scope, an interpretability toolkit using Sparse Autoencoders (SAEs) to decompose hidden representations of Qwen3 and Qwen3.5 models into interpretable features for model optimization.
FlashQLA: CP-/Bwd-Friendly Fused Linear Attention Kernels for GDN
Qwen has open-sourced FlashQLA, a high-performance linear attention kernel library built on TileLang that achieves 2-3x forward and 2x backward speedups for Gated Delta Network (GDN) layers on NVIDIA Hopper GPUs.
Qwen3.6-27B release notes / what's new
Qwen has released Qwen3.6-27B, a dense 27-billion-parameter multimodal model that outperforms the larger Qwen3.5-397B-A17B on all major agentic coding benchmarks.
Qwen3.6-Max-Preview release notes / what's new
Qwen3.6-Max-Preview is a proprietary preview model from Qwen that improves upon Qwen3.6-Plus in agentic coding, world knowledge, and instruction following.
Qwen3.6-35B-A3B Release Notes
Qwen has released Qwen3.6-35B-A3B, an open-source mixture-of-experts model with 35 billion total parameters and 3 billion active parameters that rivals larger dense models in agentic coding and multimodal reasoning.
Sandakan Central Market sign vertical text in 2024 – “2006"
The vertical text on the right side of Sandakan Central Market’s main sign in 2024 reads “2006”.
Qwen3.5-Omni release notes
Qwen announced Qwen3.5-Omni, a new omnimodal LLM that handles text, images, audio, and video, supports 256k context, 113-language speech recognition, 36-language synthesis, and adds real-time features like semantic interruption, websearch, voice control, and voice cloning.
Qwen3.5-Max-Preview Release on LMSys Arena
Qwen has deployed Qwen3.5-Max-Preview to the LMSys Arena for community evaluation ahead of its full release scheduled within two weeks.
Qwen 3.5‑397B‑A17B release: hybrid linear‑attention MoE model with 1 M token context and state‑of‑the‑art multimodal performance
Qwen 3.5‑397B‑A17B is a 397 billion‑parameter multimodal model that activates only 17 billion parameters per token, delivering state‑of‑the‑art performance on language, coding, reasoning and vision tasks while being up to 19× faster than its predecessor.
Qwen-Image-2.0 Release: Professional Infographics and Photorealism
Qwen-Image-2.0 is a unified image generation and editing model that supports 1k-token instructions for professional infographics and native 2K resolution for high-fidelity photorealism.
Qwen3-Coder-Next Release: High-Efficiency Agentic Coding Model
Qwen3-Coder-Next is an open-weight model based on a hybrid attention and MoE architecture that achieves over 70% on SWE-Bench Verified, offering performance comparable to models 10-20x larger.
Qwen3-ASR and Qwen3-ForcedAligner Release
Qwen has open-sourced Qwen3-ASR (1.7B and 0.6B) and Qwen3-ForcedAligner-0.6B, providing state-of-the-art multilingual speech recognition and non-autoregressive timestamp prediction under the Apache 2.0 license.
Qwen3-Max-Thinking Release Notes
Qwen has released Qwen3-Max-Thinking, a flagship reasoning model featuring adaptive tool-use and a multi-round test-time scaling strategy to compete with GPT-5.2-Thinking and Claude-Opus-4.5.
Qwen3-TTS Release Notes: Open-Source Voice Design, Cloning, and Generation
Qwen has open-sourced the Qwen3-TTS family, featuring 0.6B and 1.7B models that enable high-fidelity voice cloning, natural language-based voice design, and ultra-low latency streaming speech generation across 10 languages.
Qwen3-VL-Embedding and Qwen3-VL-Reranker Release
Qwen has released Qwen3-VL-Embedding and Qwen3-VL-Reranker, a series of multimodal models designed for high-precision cross-modal retrieval across text, images, screenshots, and video.
Qwen-Image-2512 release notes / what's new
Qwen-Image-2512 is a December update to the Qwen-Image text-to-image model that significantly improves human realism, natural detail rendering, and complex text layout accuracy.
Qwen-Image-Edit-2511 Release Notes
Qwen-Image-Edit-2511 is an updated image editing model that improves character consistency, integrates community LoRAs, and enhances geometric reasoning and industrial design capabilities.
Qwen3-TTS-VD-Flash and Qwen3-TTS-VC-Flash release: controllable voice design and rapid multilingual voice cloning
Qwen released Qwen3‑TTS‑VD‑Flash for natural‑language voice design and Qwen3‑TTS‑VC‑Flash for 3‑second multilingual voice cloning, both outperforming leading TTS systems on controllability and accuracy.
Qwen-Image-Layered: Layered Decomposition for Inherent Editability
Qwen-Image-Layered is a new model that decomposes images into multiple RGBA layers, allowing for independent manipulation of image components without affecting other content.
Qwen3-Omni-Flash-2025-12-01 Release Notes
Qwen3-Omni-Flash-2025-12-01 is an upgraded native multimodal model that enhances real-time audio-visual interaction, system prompt control, and multilingual support across text, image, audio, and video modalities.
SAPO: A Stable and Performant Reinforcement Learning Method for Training Large Language Models
Qwen introduces Soft Adaptive Policy Optimization (SAPO), an RL method that replaces hard clipping with a smooth, temperature-controlled gating function to improve training stability and sample efficiency in LLMs.
Qwen3-TTS-Flash Update: 49 Timbres, 10 Languages, and 9 Dialects
Qwen3-TTS-Flash is a flagship text-to-speech model supporting 49 high-quality timbres, 10 languages, and 9 Chinese dialects with improved prosody and natural speech rates.
Qwen DeepResearch 2511 Release
Qwen has released Qwen DeepResearch 2511, an AI research assistant that utilizes a multi-agent collaborative mechanism to reduce the friction between initial inspiration and deep technical validation.
Qwen3Guard Release: Real-time Safety Guardrails for LLMs
Qwen introduces Qwen3Guard, a family of safety guardrail models available in Gen and Stream variants to provide real-time, multilingual safety classification for prompts and responses.
Qwen-Image-Edit Release Notes
Qwen-Image-Edit is a 20B parameter image editing model that enables precise semantic and appearance editing, including bilingual text modification, by leveraging Qwen2.5-VL and a VAE Encoder.
Qwen-Image Release: Native Text Rendering and Precise Image Editing
Qwen-Image is a 20B MMDiT image foundation model that provides state-of-the-art complex text rendering and precise image editing capabilities.
Qwen GSPO: Scalable Reinforcement Learning for Language Models
Qwen introduces Group Sequence Policy Optimization (GSPO), a sequence-level RL algorithm that improves training stability and efficiency over GRPO, particularly for Mixture-of-Experts (MoE) models.
Qwen-MT Turbo Release Notes
Qwen has released Qwen-MT (qwen-mt-turbo), a lightweight MoE-based translation model supporting 92 languages with high customizability and low API costs.
Qwen3-Coder Release Notes
Qwen has released Qwen3-Coder, a Mixture-of-Experts model that achieves state-of-the-art open-model performance in agentic coding, browser-use, and tool-use, comparable to Claude Sonnet 4.
Qwen-TTS Update: Support for Chinese Dialects and Bilingual Synthesis
Qwen has released an update to Qwen-TTS (qwen-tts-latest) that introduces support for Pekingese, Shanghainese, and Sichuanese dialects alongside Chinese-English bilingual synthesis.
Qwen VLo: Unified Multimodal Understanding and Generation
Qwen VLo is a unified multimodal model that bridges the gap between perception and creation by combining high-quality image generation, open-ended editing, and precise visual understanding in a single system.
Qwen3 Embedding and Reranker Release
Qwen has released the Qwen3 Embedding series, a set of proprietary models based on the Qwen3 foundation model designed for state-of-the-art text embedding, retrieval, and reranking across 100+ languages.
Qwen3 Release Notes: Hybrid Thinking and Multilingual MoE Models
Qwen has released Qwen3, a family of open-weight models featuring a hybrid thinking mode for scalable reasoning and support for 119 languages.
QVQ-Max Visual Reasoning Model Release
Qwen has released QVQ-Max, a visual reasoning model capable of analyzing images and videos to solve complex problems in mathematics, programming, and creative tasks.
Qwen2.5-Omni Release: End-to-End Multimodal Model for Real-Time Interaction
Qwen has released Qwen2.5-Omni, an end-to-end multimodal model capable of processing text, images, audio, and video to generate real-time streaming text and natural speech responses.
Qwen2.5-VL-32B Release Notes
Qwen has released Qwen2.5-VL-32B-Instruct, a vision-language model optimized via reinforcement learning for superior mathematical reasoning, fine-grained image understanding, and human-aligned responses.
QwQ-32B release notes / what's new
Qwen has released QwQ-32B, a 32-billion parameter reasoning model that leverages scaled reinforcement learning to achieve performance comparable to the much larger DeepSeek-R1.
QwQ-Max-Preview Release
Qwen has introduced QwQ-Max-Preview, a preview reasoning model built on Qwen2.5-Max that excels in mathematics, coding, and Agent-related workflows.
Qwen2.5-Max Release Notes
Qwen has released Qwen2.5-Max, a large-scale Mixture-of-Experts (MoE) model pretrained on over 20 trillion tokens and available via API and Qwen Chat.