🇨🇳China·AI News·11 Nov 2026·via Tech Node

    Alibaba opens Qwen4-Max preview with a million-token context window

    Alibaba has opened a preview of Qwen4-Max, its next-generation flagship AI model. Qwen4-Max supports image and video inputs and can process up to a million tokens of context. Developers can access the model through Alibaba’s Token Plan subscription, with early use also available in the Qwen app, Qoder, QwenWork, and QwenStudio.

    Alibaba's Qwen4-Max model is characterized by its extensive context processing capability. Its reported ability to handle up to a million tokens means it can process a very large amount of information, equivalent to a substantial document or an extended conversation, within a single interaction. This capacity supports more complex and sustained tasks, where an AI needs to maintain a broad understanding of the ongoing data.

    The model also supports image and video inputs, indicating its multimodal capability. This allows Qwen4-Max to interpret and work with different forms of data beyond just text, potentially enabling applications that integrate visual and auditory information alongside textual cues.

    Developers can access the Qwen4-Max preview through Alibaba’s Token Plan subscription, suggesting a structured, usage-based access method. Additionally, the model is available for early use within Alibaba’s own applications: Qwen app, Qoder, QwenWork, and QwenStudio. These integrations offer direct engagement for users seeking to leverage the model’s capabilities within specific workflows without requiring deep technical development.

    Share this article

    Go deeper
    Original reporting by Tech NodeWe don't republish, read the full story →

    Related reading

    6 stories