Yu Jidong, Senior Vice President of iFLYTEK, revealed that the iFLYTEK Spark X1 large model is poised for a significant upgrade. Following this enhancement, the model's performance across general tasks such as reasoning, text generation, and language comprehension will be rigorously benchmarked against the industry giants, OpenAI GPT-1 and DeepSeek R1. Earlier, in January 2025, iFLYTEK introduced Spark X1, marking the debut of a deep reasoning model trained entirely using domestic computational resources. Since then, the company has relentlessly pursued optimizations through continuous research and development. iFLYTEK anticipates completing the training for the new version within the next three months, aiming to match or even surpass OpenAI GPT-1 in mathematical problem-solving and process thinking chain capabilities during comprehensive benchmarking tests.
