Ring-mini-linear-2.0

Training compute

1.7×10²³ FLOP

Parameters

16.4B

Published

Oct 23, 2025

Ring-mini-linear-2.0 is an AI model developed by Ant Group (China), first published in October 2025. It works in the language domain, on tasks such as language modeling/generation, question answering, quantitative reasoning and 2 more.

Training it took an estimated 1.7×10²³ FLOP of compute (estimation method: operation counting). The model has 16,400,000,000 parameters. It was trained on roughly 600B datapoints. Training ran on 288 NVIDIA H800 SXM5.

Access: Open weights (unrestricted). Its weights are openly available. It is built on top of Ling-mini-base-2.0-20T. Epoch AI rates the confidence of this record as confident.

Full record

Organization: Ant Group
Country of organization: China
Domain: Language
Task: Language modeling/generation, Question answering, Quantitative reasoning, Code generation, Mathematical reasoning
Training compute: 1.7×10²³ FLOP
Compute estimation method: Operation counting
Parameters: 16,400,000,000
Dataset size: 600B
Training hardware: NVIDIA H800 SXM5
Chips used: 288
Training power draw: 393.6 kW
Numerical format: FP8
Model accessibility: Open weights (unrestricted)
Open weights: Yes
Base model: Ling-mini-base-2.0-20T
Epoch confidence: Confident