Alibaba's Qwen team published Qwen-Image-2.1 on September 20, 2026: 7B parameters, native RGBA output, up to 10 reference images, and day-0 support across Diffusers, ComfyUI and SGLang. It is also the first Qwen-Image release you are not allowed to sell work from.
Alibaba released Qwen3.8-Flash-Next as open weights on Hugging Face and ModelScope, the first public preview of its next-generation Qwen4 architecture.
Swiftlet runs the real Qwen3-Next-80B on a Mac in 4.3 GB of RAM and a 35B model on an iPhone, using expert streaming to keep only active experts in memory.
Alibaba shipped Qwen-Image-3.0 on July 21, 2026, its most capable image model yet, but the first with no open weights, no benchmarks, and no technical report.
Qwen 3.7 Plus adds native vision and video understanding to Alibaba Bailian platform, challenging GPT-5.5 and Gemini for multimodal creative workflows.
NVIDIA ships an NVFP4 4-bit quantized build of Qwen3.6-35B-A3B, cutting GPU memory 3x with under 1% accuracy loss on eight benchmarks.
Qwen3.7-Max sets a new bar for non-hallucination rate among frontier agent models. Here is how it stacks up against Claude Opus 4.7, Gemini 3.1 Pro, and GPT-5.5 across the four reliability dimensions that decide which model goes into production.
Alibaba pushed Qwen 3.7 Max Preview and Qwen 3.7 Plus Preview to Arena and Qwen Chat for testing. Max sits 13th overall on Arena Text and Plus is 16th on Vision.
Alibaba launches Qwen-Image-2.0-Pro on April 25 with native 2K output, multilingual text rendering, and $0.075 per image API pricing. Arena Text-to-Image rank 9 globally.
Alibaba's Qwen team just released Qwen 3.5 Small, a series of four open-source models ranging from 0.8B to 9B parameters that run on phones, IoT devices, and browser tabs. The smallest model operates on a seven-year-old Samsung Galaxy S10E at 12 tokens per second