# Alibaba Tongyi Open-Sources Multi-Agent Development Framework AgentScope 2.0

- Category: Flash (快讯)
- Published: 2026-05-26T11:56:00.000Z
- Source date: 2026-05-26
- HTML: https://www.88z.ai/en/news/flash-20260526-c253770b69197443
- Markdown: https://www.88z.ai/en/news/flash-20260526-c253770b69197443.md
- Other language: https://www.88z.ai/news/flash-20260526-c253770b69197443.md

Alibaba Tongyi Lab releases AgentScope 2.0, featuring systematic upgrades for production stability. The new version introduces model fault tolerance, real-time message visibility, permission-based interception of high-risk operations, and context optimization to enhance multi-agent reliability in real-world scenarios.

## At a glance

- Adds model fault tolerance and permission interception to enhance production security and stability
- Refactors message module and optimizes context management for execution visibility and long-chain memory retention

## The update

Alibaba Tongyi Lab has officially launched AgentScope 2.0, focusing on stable operation in real-world scenarios and production-level integration through systematic upgrades. To address pain points in actual deployment, the new version adds a model fault tolerance mechanism that supports automatic retries and switching to backup models upon failure, ensuring service continuity. Additionally, by refactoring the message module, it achieves real-time visibility of the execution process, facilitating developer monitoring. Furthermore, the framework introduces a permission system to automatically intercept high-risk operations for security, and optimizes context management strategies to effectively prevent memory loss in long-chain interactions, comprehensively enhancing the robustness of multi-agent systems.

## Sources

- [通义实验室](https://mp.weixin.qq.com/s/AcBTESODnZqpADRELcK5MQ) (2026-05-26)

This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.
