RELEASE · MODELS · #3681
Alibaba launches Qwen4‑Max: 2.4T-parameter MoE with 1M‑token context and lower costs
Alibaba introduced Qwen4‑Max, a mixture‑of‑experts model with a 2.4 trillion‑parameter configuration that activates about 71 billion parameters per token, a one‑million‑token context window, and a revamped architecture Alibaba says reduces compute on long inputs. The company highlights large efficiency and cost gains (e.g., Artificial Analysis reports $3.27 per Intelligence Index task vs $10.02 for Qwen3.8‑Max), new “Live Delivery” agent transparency features (Qwen‑Live‑Delivery‑Plugin open‑sourced), and mixed external benchmark results versus Opus 5.5, GPT‑6, Kimi K3.1 Pro and others.
KEY POINTS
- Alibaba introduced Qwen4‑Max, a mixture‑of‑experts model with a 2.4 trillion‑parameter configuration that activates about 71 billion parameters per token, a one‑million‑token context window, and a revamped architecture Alibaba says reduces compute on long inputs.
- The company highlights large efficiency and cost gains (e.g., Artificial Analysis reports $3.27 per Intelligence Index task vs $10.02 for Qwen3.8‑Max), new “Live Delivery” agent transparency features (Qwen‑Live‑Delivery‑Plugin open‑sourced), and mixed external benchmark results versus Opus 5.5, GPT‑6, Kimi K3.1 Pro and others.
- If Alibaba's claims hold, Qwen4‑Max cuts long‑context compute and operating costs substantially and adds agent transparency via Live Delivery, which could make high‑capacity models more economical for industry and finance use cases.
WHY IT MATTERS
If Alibaba's claims hold, Qwen4‑Max cuts long‑context compute and operating costs substantially and adds agent transparency via Live Delivery, which could make high‑capacity models more economical for industry and finance use cases.