Agents-A1

Model · 35B · open weights · 1 item since Jul 2026

Part ofShanghai AI Lab

Trends

Frontier capability gets cheaper with every release

Items

Jul 3
Shanghai AI Lab's Agents-A1
An open-weight 35B MoE agent matches or beats trillion-parameter models such as Kimi-K2.6 and DeepSeek-V4-Pro on long-horizon benchmarks by scaling agent horizon (45K-token trajectories) and domain-routed on-policy distillation rather than parameters.