NewsBite analysis · AI, Programming
The Decoder
Alibaba’s open-weight Qwen3.8-Max takes on long-horizon AI tasks with 2.4 trillion parameters
Alibaba has launched Qwen3.8-Max, its most advanced language model to date, equipped with 2.4 trillion parameters. This model is designed for handling complex tasks autonomously over multi-day periods, such as writing code, reproducing research findings, and engaging in sophisticated interactions. By releasing its weights to the public, the company aims to set new benchmarks in artificial intelligence capabilities. The actions of this model include iterative coding and enhanced performance across various benchmarks, solidifying its competitive standing in the AI landscape.
Act nowConfidence: HighShort term horizon
Published 3 Aug 2026, 12:10 UTC5 min read
Visit original article ↗Lead image — The Decoder
Summary
TL;DR
- Qwen3.8-Max has 2.4 trillion parameters, enhancing AI's capability for complex tasks.
- The model can work autonomously for days, significantly improving efficiency.
- Three case studies demonstrate its coding abilities and problem-solving skills.
- Weights for Qwen3.8-Max will be available next week.
- The model competes closely with other leading AI models in benchmark tests.
Interesting facts
- Qwen3.8-Max built a command-line tool in 16 days without human intervention.
- The model improved a research paper's results by 2.7 points on a benchmark.
- It quadrupled initial capital in a simulated e-commerce scenario, outperforming human teams.