option
Home
Flash News
Content
RonaldMitchell
RonaldMitchell
May 20, 2026

Tongyi Lab released Qwen3.7-Max, a new AI Agent foundation model. It achieved top scores in evaluations and demonstrated strong long-term reasoning in tests, running for 35 hours with 1158 tool calls without interruption. The model features a novel "task-runtime-validator" design to prevent overfitting and enable cross-framework generalization. It integrates office tools via Model Context Protocol and supports multi-agent systems. Qwen3.7-Max will launch on Alibaba Cloud BaiLian, compatible with OpenAI and Anthropic APIs, advancing AI agents from theory to practical engineering deployment.

Tongyi Lab released Qwen3.7-Max, a new AI Agent foundation model. It achieved top scores in evaluations and demonstrated strong long-term reasoning in tests, running for 35 hours with 1158 tool calls without interruption. The model features a novel Tongyi Lab released Qwen3.7-Max, a new AI Agent foundation model. It achieved top scores in evaluations and demonstrated strong long-term reasoning in tests, running for 35 hours with 1158 tool calls without interruption. The model features a novel
Comments (0)
0/300
OR