AI模型精选73°

StartLux 27B本地模型在中国AI基准测试中超越DeepSeek V4 Flash

StartLux's 27B Local Model Beats DeepSeek V4 Flash in China AI Benchmark

精选理由

StartLux发布了27B本地模型,在中国基准测试中击败了DeepSeek V4 Flash,性能接近万亿参数级别。

AI 摘要

StartLux的270亿参数本地模型在CAICT MCP测试中排名第二,性能超过DeepSeek-V4-Flash。该模型达到了万亿参数级别的能力表现。CAICT MCP测试是中国AI领域的重要基准。

原文 · Pandaily

StartLux's 27B Local Model Beats DeepSeek V4 Flash in China AI Benchmark

StartLux's 27-billion-parameter local model took second in the CAICT MCP test, outpacing DeepSeek-V4-Flash and reaching trillion-parameter capability territory.