Subtle announces "Voicebuds" AI earphones with "voice isolation" just before CES—enhancing calls and transcription for $199
While the era of interacting with AI via voice is advancing, the real-world barriers are "noise" and "the eyes of others." Cafes, offices, trains—there are surprisingly few "environments where you can talk." That is why a startup has emerged that focuses not on the music experience, but on "isolating your voice."The wireless earphones (Voicebuds) announced by voice AI startup Subtle are making their move just before CES, marketing themselves on call clarity and high-precision transcription for voice memos.
1. From "noise canceling" to "voice isolation": Subtle's goal
Subtle emphasizes not general noise canceling (eliminating ambient sound), but "voice isolation" for voice input. It makes it easier for computers to understand your speech even in noisy environments—in other words, the protagonist is not "listening," but "input."
The product is scheduled to ship in the U.S. within the next few months, priced at $199. It comes with a one-year subscription to an iOS/Mac app and is designed to allow for voice memos and AI chat without key operations. Furthermore, they explain that it is equipped with a custom chip that can "wake up the iPhone even when it is locked."
2. What can it do? "Dictation anywhere" is the biggest differentiator
2-1. Speaking in "any app" = going outside the dictionary app
Subtle explicitly states that it competes with "AI dictation app groups" such as Wispr Flow, Willow, Monologue, and Superwhisper, and puts the point that "you can use dictation input in any app with Voicebuds" at the forefront.
This is important; in 2025, when dictation apps have become widespread, "accuracy improvement + automatic formatting" has become the norm, and "where and how you can use it" has become the deciding factor.
2-2. The demo showed the strength of "noise" and "whispering"
In a demo seen by TechCrunch, it is said that the device successfully picked up audio even in noisy backgrounds and succeeded in transcribing text for voice memos even in situations where co-founder and CEO Tyler Chen was whispering.
Subtle further claims that it has one-fifth the errors (5 times fewer) than "AirPods Pro 3 + OpenAI's transcription model." However, this is currently a "manufacturer's claim," and third-party verification is awaited.
3. Why earphones now? To bring the "voice UI" into daily life
Chen told TechCrunch the following:
"Voice can do more things more naturally than a keyboard. But in places where others are present, voice UI is almost never used. That is why we are delivering the experience in the form of earphones with a noise isolation model."
In short, the recognition is that what hinders the spread of voice UI is not just model performance, but the "psychological cost of the moment you use it."
Just as ring-type memo devices (Sandbar and Pebble) have tried to redesign "writing/saving" over the past year, Subtle is trying to package "speaking/inputting" into one (dictation + AI chat + voice memos).
4. Business foundation: Fundraising, partnerships, and the battleground at CES
Subtle has raised a total of approximately $6 million and has collaborated with consumer-facing companies like Qualcomm and Nothing on the deployment of noise isolation models.
Also, through selection for a Qualcomm program, it has been indicated that the company's technology could lead to OEM deployment for devices equipped with Qualcomm chips.
In the official announcement, it is stated that Voicebuds will be exhibited at CES's Eureka Park, and they will also conduct a "whisper-level" demo in a crowded environment. There are two colors: Parchment White and Ink Black, with shipping scheduled for early 2026 in the U.S., and reservations can be made via the official website.