An app to hear the world's voices

听世 Ariya

听见每一种声音

Hear every voice

实时语音识别 · 多语互译 · 跨语言对话 Speak. Listen. Understand — instantly.

即将上线 · 公测申请中 Coming soon · join the early access

Three pillars

听见 · 翻译 · 对话 Listen · Translate · Converse

三项能力,贯穿一次跨语言对话的全程。 Three capabilities. One uninterrupted cross-language conversation.

01

听见

Listen

本地 + 云端混合 ASR,支持 50+ 语言。 Hybrid on-device + cloud ASR, 50+ languages.

powered by Whisper · 50 langs

02

翻译

Translate

实时多向翻译,保留语气与停顿。 Real-time translation that keeps the tone and pauses.

双向 · 流式 · < 400 ms 延迟Bidirectional · streaming · < 400 ms

03

对话

Converse

双向声波视图,把对话变成可看见的合奏。 Two-voice waveform — duet, not duel.

声纹保留 · 端到端加密Voiceprint preserved · end-to-end

"Every voice, in earshot."

万语千声,皆在耳畔。

Showcase

听见对方的声音,也让对方听懂你的

Hear them, and let them hear you.

(示意 / Visual mock) (Visual mock)

让世界听见你,也让你听懂世界

Be heard. Hear back.

加入早鸟Get early access

FAQ

常见问题 Frequently asked

什么时候上线? When will it launch?

我们正在做最后的内测,预计 2026 年内开放公测。订阅"加入早鸟"在第一时间获得邀请。 Final beta in progress — public access expected within 2026. Subscribe to early access to be first in line.

支持哪些语言? Which languages?

首发支持中、英、日、韩、法、德、西、葡、意、阿、印地、越、泰等 50+ 语种,持续扩展。 Launch with 50+ languages including ZH/EN/JA/KO/FR/DE/ES/PT/IT/AR/HI/VI/TH; growing.

离线可用吗? Does it work offline?

基础语种支持端侧识别,弱网或离线可继续使用;翻译质量在云端最高。 Core languages run on-device for offline use; cloud unlocks the highest translation quality.

我的语音数据如何处理? How is my voice data handled?

语音不留存,翻译过程端到端,可在设置中关闭任何云上传。 Voice is not stored. Pipelines are end-to-end; cloud upload can be turned off in settings.