Moreover, both DeepSeek-V2 Chat (SFT) and DeepSeek-V2 Chat (RL) outperform GPT-4-0613 and ERNIEBot 4.0, solidifying the position of our models in the top-tier LLMs that support Chinese. Specifically, DeepSeek-V2 the reasoning capability of DeepSeek-V2 Chat (RL) still lags behind giant models, such as Erniebot-4.0 and GPT-4s. 19 Model Overall Reasoning 中文推理 Language 中文语言 Avg. Math. Logi. Avg. Fund. Chi. Open 67 8.47 8.65 DeepSeek-V2 Chat (RL) 7.91 7.45 7.77 7.14 8.36 8.10 8.28 8.37 8.53 8.33 8.53 ERNIEBot-4.0-202404*（文心一言） 7.89 7.61 7.81 7.41 8.17 7.56 8.53 8.13 8.45 8.24 8.09 DeepSeek-V2 Chat (SFT) 7.74 7