88Z AI, 7 August — Ant Bailin has released Ling-3.0-tiny, a native hybrid reasoning model with 7.9B total parameters and 1.3B activated per token. Optimized for mathematical reasoning and edge deployment, it supports tool calling and fully local offline inference.
At a glance
- The model has 7.9B total parameters with only 1.3B activated per token, optimized for math reasoning and edge deployment
- It natively supports tool calling for UI automation and enables fully local offline inference
The update
Ant Bailin has officially launched Ling-3.0-tiny, a native hybrid reasoning model. With a total parameter count of 7.9B, it activates only 1.3B parameters per token during operation. This architectural design significantly reduces computational resource consumption while maintaining high performance. The model is deeply optimized for mathematical reasoning tasks and edge device deployment, aiming to enhance operational efficiency in resource-constrained environments. Furthermore, Ling-3.0-tiny natively supports tool calling, enabling it to directly drive browser user interfaces and automate operations on mobile devices, thereby expanding its interactive capabilities in practical application scenarios. Notably, the model supports fully local offline inference, allowing it to run independently on local devices without relying on cloud connections, which offers users greater data privacy protection and deployment flexibility.
Sources & editorial note
This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.
- Ling-3.0-tinySource published
Source dates above refer to the original announcements; the date under the headline is this article’s publication on 88Z AI.