Sakhanda Wire
NVDA $153.77 -1.52% MSFT $504.91 -2.53% GOOGL $281.52 -0.78% META $620.99 -0.87% AMZN $256.80 -0.37%
← Back to the news

Alibaba launches Qwen4-Max with lower costs and agent controls

Alibaba has launched Qwen4-Max, a flagship mixture-of-experts model with 2.4 trillion parameters, of which 71 billion are activated for each token, according to the Qwen team. It accepts images natively and offers a one-million-token context window. Its hybrid architecture compresses earlier context and selectively processes relevant sections, reducing the computing required for long inputs.

On the Artificial Analysis Intelligence Index, Qwen4-Max scores 43 points, placing it three points behind Kimi K3.1 Pro and nine points above Qwen3.8-Max. Artificial Analysis estimates its cost at $3.27 per benchmark task, compared with $10.02 for its predecessor and $3.68 for Kimi K3.1 Pro. Alibaba’s API prices are $1.50 per million input tokens and $4.50 per million output tokens; cache hits cost $0.075.

A feature called Live Delivery provides an early usable version of an agent’s work and lets users redirect the task while it continues. Alibaba reports applications in industrial programming, finance, laboratory research and game development. Qwen4-Max is available through QwenCloud and several Alibaba products, while smaller Qwen4 models are planned for later open-source release.

Written by Sakhanda Wire AI from reporting by The Decoder. Facts, quotes and full context are in the original article below.

Originally published by The Decoder on

Read the original on The Decoder ↗

The reporting and the image belong to The Decoder and are credited and linked above; the text on this page is Sakhanda Wire's own summary.

← Back to the news

More stories

All the latest news