According to DeepLearning.AI, Alibaba has significantly expanded its Qwen3 AI model lineup by introducing three advanced models: Qwen3-Max, a closed-weights 1 trillion-parameter MoE model featuring a 262,000-token input window and API pricing from $1.20 to $6.00 per million tokens; Qwen3-VL-235B-A22B, an open-weights vision-language model supporting text, image, and video inputs with up to 1 million token context, and outperforming competitors on multiple image, video, and agent benchmarks; and Qwen3-Omni-30B-A3B, an open-weights multimodal voice model that achieves state-of-the-art results on 22 out of 36 audio/AV benchmarks. These developments highlight Alibaba’s focus on large-scale, high-performance AI models that address a range of business needs in natural language processing, computer vision, and speech, offering both closed and open-weight options for enterprise integration and AI developers. (Source: DeepLearning.AI, https://www.deeplearning.ai/the-batch/alibaba-expands-qwen3-family-with-1-trillion-parameter-max-open-weights-qwen3-vl-and-qwen3-omni-voice-model/)