{
  "id": 20305001,
  "title": "With Qwen4-Max, Alibaba puts the open source jacket back on the hanger",
  "url": "https://urgent.news/2026/11/16/with-qwen4-max-alibaba-puts-the-open-source-jacket-back-on-the-hanger",
  "topic": "tech",
  "section": "Tech",
  "published": "2026-11-16T16:54:17.000Z",
  "source": {
    "name": "The New Stack",
    "slug": "the-new-stack",
    "url": "https://thenewstack.io/qwen4-max-alibaba-open-weights/"
  },
  "original_language": "en",
  "account": "Alibaba has officially introduced Qwen4-Max, the latest model in its advanced Qwen4 series, designed for tackling complex professional tasks and autonomous AI agents. The company teased the new Qwen4 family last week, promising enhancements in coding, reasoning, and real-world knowledge, initially releasing Qwen4-Max in preview through various products. Now, Alibaba has officially detailed Qwen4-Max's capabilities, positioning it as an always-available \"AI teammate\" for software development, scientific research, and beyond.\n\nBuilt on a new MoE architecture with native vision capabilities and a massive 1M-token context window, Qwen4-Max boasts significant improvements over its predecessor, Qwen3.8-Max. Alibaba claims up to 50% fewer tokens while reducing the average cost per task, resulting in a more efficient and cost-effective AI solution. The model also supports images alongside text and comes with a one-million-token context window, enabling it to process extensive information in a single interaction.\n\nAvailable through Alibaba's QwenCloud platform, Qwen4-Max can be accessed via API pricing starting at $1.50 per million input tokens and $4.50 per million output tokens. It is also integrated into QwenWork, Qoder, QwenStudio, and Vercel's AI Gateway, allowing developers to leverage the model through coding tools like Claude Code, Codex, and Cursor. Alibaba reports substantial improvements over Qwen3.8-Max, particularly in complex tasks involving multiple tools and extended autonomous execution periods. On an internal benchmark called Work360, Qwen4-Max scored 52.69 compared to 47.74 for Qwen3.8-Max, with between 24% and 35% fewer tokens on average.",
  "summary": "Alibaba has officially unveiled Qwen4-Max, the first model in its next-generation Qwen4 family, with a focus on complex professional tasks The post With Qwen4-Max, Alibaba puts the open source jacket back on the hanger appeared first on The New Stack .",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}