In a blind testing list, LMSYS Yi-Lightning surpassed OpenAI's GPT-4o-2024-05-13 and Anthropic's Claude 3.5 Sonnet, achieving sixth place globally and first in China.
01.AI announced that in the global competition of models, LMSYS's large-scale Chinese model has achieved the best performance for the first time, outshining OpenAI's GPT-4o.
In addition to enhancing model performance, 01.AI claims that the output speed of Yi-Lightning has also increased. The time to receive the first packet has been halved compared to Yi-Large, which was released in the first half of this year, and the maximum generation speed has risen by nearly 40%. Furthermore, Yi-Lightning has significantly reduced the costs associated with obtaining responses.