大语言模型列表

O4-mini is OpenAI's high-efficiency reasoning model launched in 2025. It delivers strong logical reasoning, tool use, and vision capabilities at a fraction of the cost and latency of larger models.

DeepSeek-V3 is DeepSeek's latest open-source large language model, featuring a 671B Mixture of Experts (MoE) architecture with 37B activated parameters.