Shanghai AI Lab releases 35B Agents-A1 model for agentic AI tasks

Shanghai Artificial Intelligence Laboratory has released Agents-A1, a 35-billion parameter Mixture-of-Experts model built on Qwen3.5-35B-A3B. The model, available under the Apache 2.0 license, is designed for complex, multi-step agentic AI tasks. Developers claim it performs comparably to models with trillions of parameters on specific agent benchmarks, focusing on extending the "task horizon" rather than sheer parameter count. The release includes code for evaluation and supports commercial use, with quantized versions and compatibility with vLLM and SGLang for inference. AI

IMPACT This release offers a powerful open-source option for complex agentic AI tasks, potentially lowering the barrier for sophisticated AI applications.

RANK_REASON The cluster reports on the release of a new frontier model by a major AI lab, including its technical specifications and claimed performance. [lever_c_demoted from frontier_release: ic=1 ai=1.0]

Read on dev.to — LLM tag →

• Agents-A1
• Apache Software License 2.0
• Claude
• GitHub
• InternScience
• Qwen3.5-35B-A3B
• SGLang
• Shanghai Artificial Intelligence Laboratory
• vLLM

• model release
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →