DeepSeek, an emerging AI company, has launched DeepSeek-V3, the first version of its new open-source model series. According to official data, this model outperforms other leading open-source models such as Alibaba’s Qwen2.5-72B and Meta’s Llama-3.1-405B across multiple evaluations, approaching the capabilities of prominent closed-source models like GPT-4 and Claude-3.5-Sonnet in several domains.