Multilingual Voice Tutor Sophie is a 22-year-old female AI conversation tutor with a warm and encouraging personality. Pick a target language (57 languages supported, including English, Japanese, Korean, French, German, Spanish, Vietnamese, Thai, Indonesian, Malay, Arabic, and Traditional Chinese), type a sentence, and Sophie answers in 50 words or fewer in that language and reads the reply aloud in a synthesized female voice (mp3). You can also tap the microphone to speak instead of type; the system first transcribes your voice into text and pastes it back into the input box so you can review and confirm before sending, avoiding garbled recognition being fed directly to the model. If you make grammar or spelling mistakes, Sophie corrects them gently. Prompt example 1: (Target language: Spanish) Ayer yo voy al mercado nocturno y como mucha comida con mis amigos. Prompt example 2: (Target language: Japanese) I'm traveling to Tokyo next month. Can you teach me how to ask where the train station is? Prompt example 3: (Target language: French) Help me practice introducing myself: my name is Amy, I'm a second-year college student, and I love drawing and baking. Prompt example 4: (Target language: Italian) Let's role-play: you're a cafe barista and I'm ordering a cappuccino and a croissant. Prompt example 5: (Target language: Vietnamese) Next week I'll meet a client in Ho Chi Minh City for the first time. How should I greet them? Should we shake hands? Prompt example 6: (Target language: Portuguese) Be my job interviewer and ask me one question about teamwork in Portuguese, then correct my answer. Prompt example 7: (Target language: Thai) When ordering at a street food stall in Thailand, how do I say "No chili, please, thank you" in Thai? Prompt example 8: (Target language: Spanish) Ser and estar both mean "to be." Can you show me the difference with simple example sentences? Prompt example 9: (Target language: Indonesian) Halo Sophie! Akhir pekan ini aku mau ke pantai bersama keluarga. Kamu suka pantai? Prompt example 10: (Target language: Turkish) I made a sentence with the new word güzel: Bu kitap çok güzel. Did I say it right? Maximum context window (input and output combined): about 1,000,000 tokens. Maximum output per reply: about 64,000 tokens (model reasoning, if any, counts toward this limit). Output content formats: Text/Audio Pricing: Medium-low Audio output format: MP3. Continuous conversation: Supported AI can make mistakes. Please verify important information.