Encontrado(s) 7 {posts} sobre este tópico.
Standard mode is the one most of you use every day. Some unusual individual subtitles used to trip up the AI and drag their neighbors down with them. We strengthened how Standard mode handles those cases — and we will keep improving it. No app update needed.
Due to an intermittent subtitle timeline issue in server-side Advanced mode caused by new model instability, we are urgently working on a fix. Until then, all Advanced mode requests are routed to Standard mode, billed at Standard pricing.
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.
SubEnvoy v1.2.1 is officially released for both macOS and iOS. Discover the new Pro models on Mac and significant performance optimizations for mobile.
Pre-release deep dive into SubEnvoy v1.2.1 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.
