paper: https://t.co/NH8TYhD0Yf

Paper page - Ling and Ring 2.6 Technical Report: Efficient and Instant Agentic Intelligence at Trillion-Parameter Scale

Source: https://huggingface.co/papers/2606.15079 Authors:

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

Abstract

Ling-2.6 and Ring-2.6 models are presented as scalable solutions for agentic intelligence, featuring architectural upgrades and specialized training methods to balance fast response times with advanced reasoning capabilities.

Efficient and scalable agentic intelligence requires models that can deliver both low-latency responses and strong reasoning capabilities while remaining practical to train, serve, and deploy. In this report, we present Ling-2.6 and Ring-2.6, a family of models designed to address this challenge at scale. Ling-2.6 is optimized for instant response generation and high capability per output token, whereas Ring-2.6 is tailored for deeper reasoning and more advanced agentic workflows. Instead of training from scratch, we upgrade the Ling-2.0 base model througharchitectural migration pre-trainingand large-scale post-training. This upgrade is guided by a unified co-design of model architecture, optimization objectives, serving systems, and agent training environments, enabling improvements in both model capability and deployment efficiency. At the architectural level, we introduce a hybrid linear attention design that integratesLightning AttentionwithMLA, improving the efficiency of long-context training and decoding. To further enhance token efficiency, we optimize capability per output token throughEvolutionary Chain-of-Thought,Linguistic Unit Policy Optimization,bidirectional preference alignment, andshortest-correct-response distillation. For agentic capabilities, we proposeKPop, areinforcement learning frameworkdesigned to support stable training of Ring-2.6-1T on large-scale environment-grounded data.KPopimproves training efficiency throughasynchronous schedulingacross coding, search, tool use, and workflow execution, enabling scalable learning from complexagent-environment interactions. Together, Ling-2.6 and Ring-2.6 provide a practical pathway toward efficient, scalable, and open agentic systems. We open-source all checkpoints in the 2.6 family to support further research and development in practical agentic intelligence.

View arXiv pageView PDFAdd to collection

Models citing this paper5

#### inclusionAI/Ling-2.6-flash Text Generation• 107B• Updated7 days ago • 10.8k • 497 #### inclusionAI/Ling-2.6-1T Text Generation• 1T• Updated7 days ago • 538 • 472 #### inclusionAI/Ring-2.6-1T Text Generation• 1T• Updated7 days ago • 1.34k • 102 #### inclusionAI/Ling-2.6-1T-base Text Generation• 1T• Updated1 day ago • 229 • 13 Browse 5 models citing this paper## Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2606.15079 in a dataset README.md to link it from this page.

Spaces citing this paper19

Browse 19 spaces citing this paper## Collections including this paper3