"iFlytek releases iFlytek Star Fire 3.0, comprehensive ability surpasses ChatGPT"

The Keda Xunfei Global 1024 Developer Festival was held in Hefei, where Liu Qingfeng, the chairman of Keda Xunfei, unveiled Xunfei Spark 3.0, claiming that its comprehensive capabilities have surpassed ChatGPT.

Xunfei Spark 3.0 is a cognitive large-scale model that achieves continuous improvement in seven major areas, including text generation, language understanding, knowledge questioning, logical reasoning, mathematical ability, coding ability, and multimodal ability. According to the official statement, Xunfei Spark 3.0's Chinese language ability has surpassed ChatGPT, while its English language ability is comparable. Liu Qingfeng revealed that Xunfei Spark 3.0 is far ahead of ChatGPT in terms of mathematical ability and coding ability.

Liu Qingfeng pointed out that cognitive large-scale models are ushering in a new era of growth, and large-scale models are becoming increasingly relevant to everyone. Liu Qingfeng stated, "In the current global political, economic, and military challenges and uncertainties, general artificial intelligence is the most certain source of growth in this era."

Keda Xunfei stated that with the help of the Xunfei iFlyCode tool, the development work that originally took three months for programmers to complete platform migration was completed in just one month. In addition, Keda Xunfei announced that since the release of the intelligent programming assistant iFlyCode based on the Spark code large-scale model on August 15th, it has been deeply integrated with 107 institutions such as JD Cloud and Softtek. In terms of internal efficiency improvement, based on the internal usage data of Keda Xunfei, the upgraded iFlyCode V2.0 has a wider range of efficiency improvement, achieving 50% efficiency improvement in the design stage, and 37% and 44% efficiency improvement in the coding stage and testing stage, respectively.