在「4」話題下找到 {posts} 篇文章。
SubEnvoy spent the last stretch not on new features, but on the unglamorous work of tuning its translation engine: translating the same subtitles over and over, comparing the results, and adjusting until the output truly fit subtitle translation.
SubEnvoy has moved Standard mode to DeepSeek V4 Flash 0731 and Advanced mode from Gemini to Qwen 3.7 Plus as we reduce our reliance on restricted closed commercial models.
SubEnvoy's translation engine gets a significant upgrade — DeepSeek V4 Flash powers Standard mode for faster everyday translations, while Advanced mode moves to Gemini 3.1 Flash Lite for deeper context understanding. Both models graduate from preview to production-tier stability.
Pre-release deep dive into SubEnvoy 1.2.1 for macOS, featuring the new multi-model switching capability. Learn the technical differences between Lite and Pro models and why iOS performance constraints shaped our strategy.
