Translated data: Chairman Liu Qingfeng of iFlytek announced the "iFlytek Spark" Cognitive Large Model V3.0, which is set to benchmark against GPT-4 in the first half of 2024. The new model has achieved comprehensive improvements in seven dimensions including text generation and language understanding, with its English capabilities benchmarking against ChatGPT. iFlytek emphasizes that domestic large models need to enhance in multiple aspects to achieve true superiority.