inclusionAI/Ling-3.1-flash-1019 · Hugging Face
retained text
from inclusionAI: We introduce **Ling-3.1-flash-1019**, an upgraded version of [Ling-3.1-flash](https://huggingface.co/inclusionAI/Ling-3.1-flash), with **560B** total parameters, **25B** active parameters per token, and a **1M-token context window**. With further post-training optimizations, **Ling-3.1-flash-1019** demonstrates significant improvements in reasoning and real-world agentic capabilities. Its **AA-Briefcase** score rises from approximately 1400 to **1524 Elo**, while performance on **AutomationBench-AA** increases from 62% to **68%**. The model also makes gains on scientific benchmarks, including **CritPt** (18% → **26%**) and **SciCode** (54% → **57%**). Overall, it scores **44** on the **Artificial Analysis Intelligence Index v4.3.2**, up from 41 for its predecessor. These gains highlight the model's improved capabilities in complex reasoning, multi-step task execution, and agentic workflows across diverse real-world scenarios. [https://huggingface.co/inclusionAI/Ling-3.1-flash-base](https://huggingface.co/inclusionAI/Ling-3.1-flash-base) We introduce **Ling-3.1-flash-base**, the base checkpoint of [Ling-3.1-flash](https://huggingface.co/inclusionAI/Ling-3.1-flash), released to support research and community-driven innovation. With **560B total parameters** and **25B active parameters per token**, the model provides a foundation for continued pretraining, domain adaptation, and post-training research.
document fetched 2026-10-28 00:42 · canonical https://reddit.com/r/LocalLLaMA/comments/1xgpo4q/inclusionailing31flash1019_hugging_face