JT-MINI is a compact, high-performance language model developed by China Mobile as part of its Jiutian (九天) AI foundation model series. Released in April 2026, the model is engineered to provide efficient generative AI capabilities optimized for low-latency, terminal-side deployment and enterprise-level industrial applications. It serves as a core component of the company's "AI Plus" strategy, which aims to integrate artificial intelligence across telecommunications, energy, and manufacturing sectors.
The model features a robust 128,000-token context window, enabling it to process extensive datasets, analyze lengthy technical documentation, and maintain coherence in prolonged multi-turn interactions. Despite its smaller scale compared to trillion-parameter flagship models, JT-MINI has demonstrated high efficiency in coding and general intelligence tasks, frequently achieving top-tier scores on benchmarks like the Artificial Analysis Intelligence Index for its size class.
JT-MINI belongs to the Jiutian 3.0 ecosystem, which prioritizes architectural safety, domestic hardware compatibility, and industrial self-reliance. While China Mobile maintains an open-weights repository for specific research checkpoints, JT-MINI is primarily deployed via the Jiutian aggregation platform as a proprietary solution for high-volume processing and cost-sensitive commercial projects. It is designed to minimize hallucination in professional scenarios and offers specialized performance in domains such as code generation and software development.