正文 · AI 翻译
Thanks to IT Home reader 软媒用户1773925 for the tip!
IT Home reported on November 6 that the Ant Ling official team announced today the launch of its native multimodal model Ling-3.2-flash, with 560B total parameters, 25B active parameters, and a 1M context window.
The model natively supports image, text, and video input, with a focus on strengthening four capabilities: complex visual understanding, long-range multimodal agents, 3D understanding and interaction, and medical multimodality. It can be tried online through a chat page, supports API calls compatible with OpenAI and Anthropic protocols, and is available for a limited-time free trial for the first 2 weeks after launch.
In evaluations released at launch, the model outperformed Opus 5 (max):
- OmniDocBench1.5: 92.21 vs 90.56
- ZEROBench (Sub): 43.71 vs 39.52
- Vision2Web: 85.86 vs 80.20