没想到 Claude 还真有中文税。AI 研究员 Aran,对各大模型进行了测试,它将相同内容的文章,让 AI 翻译成不同语言,结果翻译中文时,Claude 要多花 65% 的 Token。
而其它模型,如 Gemini、Qwen、DeepSeek、Kimi,消耗量就正常很多,即使 OpenAI 的 ChatGPT 也只是多 15%。
不过好消息是,Claude 也对其它语言多收税,其中日韩收的最狠,韩国整整要多花 159%。
显示更多
Follow-up on non-English token-inefficiency with more model-language pairs:
- Chinese is cheaper than English on major Chinese models
- Gemini and Qwen provide least non-English tax
- Anthropic has the highest tax by far; Kimi is next
- Hindi is the worst-covered language here, despite its massive speaker base
显示更多