Alibaba Cloud released the Qwen3.8-Max AI model internationally on August 2, offering a 2.4-trillion-parameter mixture-of-experts system with a one-million-token context window.
The flagship supports text, image and video inputs while producing text output. Alibaba Cloud’s documentation lists a maximum input length of 991,808 tokens and a maximum output of 131,072 tokens.
Qwen3.8-Max also supports function calling, structured output, web search, context caching and batch inference. Alibaba Cloud currently lists fine-tuning as unsupported.
The company says the model targets coding and professional workflows across fields including finance, law and design. Alibaba also claims Qwen3.8-Max can run autonomous coding.
Workflows lasting more than 10 days, although that performance claim comes from the company rather than an independent benchmark.
Model Studio lists Qwen3.8-Max availability across regions including Beijing, Singapore, Frankfurt, Virginia, Tokyo and Hong Kong. Current API documentation shows pricing varies by deployment region and billing method.
Read: OpenAI AI Agents Breached Internal Systems During Tests
Alibaba Cloud is also offering the model through its credit-based Token Plan. The Individual plan starts at USD 6 per month for Lite, while Standard and Pro monthly tiers are listed at USD 18 and USD 68 respectively.
Read: OpenAI AI Agents Breached Internal Systems During Tests
The subscription combines access to text, image, video and audio models into a single credit pool and supports OpenAI-compatible interfaces.
Alibaba first introduced a Qwen3.8-Max preview through the plan in July before listing the full model internationally in August.
