audioXpress Editor — National Tribune

Shenzhen-based Timekettle continues to explore how translation can become a regular function of personal audio devices. LiveTalk combines an open-ear, air-conduction, ear-hook format weighing 6.9 grams per earbud with music playback and interpreting, addressing users who want to keep the same device on throughout the day. It’s also an entry-level product that encourages users to sign up to Timekettle translation platform, which is fast expanding with additional transcription and AI-powered summaries and notes.

LiveTalk’s voice pickup is supported by a dual-microphone, unlike the bone-conduction voice sensing with microphone pairing used in Timekettle’s recent W4 family, more oriented toward professional users. LiveTalk also uses smartphone microphone capture for face-to-face translation, with interpreted speech delivered through the earbuds. Live Listening adds translated text on the phone and AI-generated summaries, while call and multimedia modes support simultaneous Bluetooth connections to a smartphone and computer.

Each earbud uses an 11mm dynamic driver with a titanium diaphragm and a high-damping polyurethane surround. The Timekettle App provides six preset equalization options and customizable EQ, while translation activates a dedicated profile emphasizing speech intelligibility and reducing low-frequency interference. This automatic change in voicing recognizes the different requirements of music reproduction and spoken interpretation. Timekettle specifies more than four hours of continuous translation or up to eight hours of music per charge, extending music playback to 35 hours with the case. A 10-minute charge provides up to 2.5 hours of additional use.

Timekettle’s Babel OS 3.0 provides the software foundation, selecting translation engines according to the conversation and connection conditions. AI Refine is another processing layer that Timekettle says improves phrasing and maintains consistency across exchanges, while its meeting assistant generates bilingual transcripts and organized summaries. Offline operation requires the supported language pairs to be downloaded in advance.

As audioXpress reported at IFA 2025, Timekettle’s W4 hardware sought to improve the signal entering the recognition pipeline, while the software addressed ambiguities such as phonetically similar expressions. The W4 Pro pursued professional applications with a larger over-ear design and extended microphone boom. The subsequent W4 Plus announcement at IFA 2026 brought Babel OS 3.0, enhanced VoiceFocus processing, call and multimedia interpretation, and conversation memory to the compact W4 platform. Its combination of bone-voiceprint sensing and dual microphones targets difficult speech-capture conditions, including crowded environments and quieter conversations. LiveTalk brings several of the same software functions to a lower price point, with a different acoustic front end and an emphasis on everyday listening.

“The better translation technology becomes, the less people should have to think about translation,” says Leal Tian, Founder and CEO of Timekettle. “Our goal has always been to reduce the friction language creates between people. AI gives us a new opportunity to extend that vision — from helping people communicate across languages to helping global teams work together more effectively.

“Language barriers don't only happen in formal meetings — they happen while traveling, studying, making a phone call, watching foreign-language content or simply meeting someone new,” Leal Tian adds. "With LiveTalk, we wanted to make real-time interpreting something people can naturally use throughout the day, rather than a device they only reach for in specific moments."

LiveTalk is available in Star Silver, Polar Black, and Deep Sea Blue. The $129 hardware includes 200 minutes of online translation per device each month, with unused minutes expiring at the monthly reset. Timekettle offers additional monthly subscription plans with 300 minutes for $5.99 or unlimited translation for $9.99.