Subtle Voicebuds launch with AI-powered transcription that works in whispers and loud spaces

Subtle releases ear buds with its noise cancelation models | TechCrunch

Voice AI startup Subtle, which creates voice isolation models to have computers understand you better in loud environments, today launched a new pair of wireless earbuds that help users sound clear in calls and get clear transcription for notes. The company unveiled these earbuds ahead of the Consumer Electronics Show (CES) in Las Vegas and said that it plans to ship them in the U.S. in the next few months. The buds cost $199 and will come with a year-long subscription to the iOS and Mac app. The app will let users take voice notes or chat with AI without pressing any keys. The company said it is using a chip that allows it to wake the iPhone while it is locked. The startup is also trying to compete with AI-powered voice dictation apps such as Wispr Flow, Willow, Monolouge, and Superwhisper by allowing users to dictate in any app using the voice buds. The company claimed that buds would deliver five times fewer errors than AirPods Pro 3 combined with OpenAI's transcription model. In a demo seen by TechCrunch, the voice buds were able to capture audio in a noisy background. The buds also managed to capture the text for a voice note when Subtle's co-founder and CEO, Tyler Chen, was speaking in a whispering tone. "We are seeing that there is a huge move towards voice as a new interface that a lot of folks are adopting. You can do much more with voice in a natural way than with a keyboard. However, we saw that voice is rarely an interface people use when others are around. So that using our noise isolation model, we will give consumers a way to experience a voice interface in the form of our earbuds," Chen told TechCrunch over a call. Last year, companies like Sandbar and Pebble launched rings for note-taking. Chen said that with its buds combined with app, it wants to provide functionalities of different tools like dictation, AI chat, and voice notes in one package. Users can place a pre-order of these buds using the startup's site. The Voicebuds are available in black and white colorways. Subtle has raised $6 million in funding to date, and has been working with consumer companies like Qualcomm and Nothing to deploy their models for noise isolation.

The Subtle Voicebuds use AI to transcribe your words below a whisper, or in very loud spaces (like the CES show floor)

There's a good chance you spend more time talking to your phone's virtual assistant, or dictating text with your voice, instead of actually calling people these days. But, as convenient as voice input can be, you don't want to be the obnoxious person shouting commands to Siri or ChatGPT in a quiet library. And you probably won't have much luck dictating an email in a room with toddlers screaming and Peppa Pig blaring on the TV. (Ask me how I know.) At CES 2026, the startup Subtle is unveiling its solution to those issues: Voicebuds. While they look like a typical pair of wireless earbuds, they feature a custom AI model that lets you dictate text below the sound of a whisper. Additionally, they can also filter out noisy environments so that you don't have to shout for dictation and voice commands. The Voicebuds seem similar to the WHSP ring we saw at CES 2024, which let whisper to your phone, and they also compete with other AI earbuds like the Notebuds One. Subtle claims its Voicebuds deliver five times fewer transcription errors than the AirPods Pro 3 with OpenAI transcription. At $199, they're also priced competitively with Apple's best buds. As with many new hardware products, though, there's also an additional subscription for premium features on Subtle's app, including instant dictation and the ability to transcribe notes without looking at your phone. You'll get a year's worth of access to the Subtle iOS app (there's no Android support so far) when you buy the Voicebuds, but after that it's a $17 a month fee. Without the subscription, the Voicebuds still offer better overall transcription accuracy with its on-device machine learning model. I haven't been able to try the Voicebuds in person yet, but in a remote demo I watched Subtle CEO Tyler Chen use them to accurately dictate text while playing loud music. They also worked while while he was barely whispering a few sentences. (He was so quiet, I couldn't even hear what he was saying over video chat). As you'd expect, you can use Voicebuds to take calls and listen to audio just like any other pair of headphones. They also offer other modern features, like active noise cancellation and a multi-mic array. Still, I can't imagine a fledgling startup will be able to match the refined audio quality we've seen from the AirPods Pro 3 and Sony's latest buds, or their well-honed ANC capabilities. As someone who's practically attached to my AirPods Pros, primarily because they handle Siri voice commands so well, it would have to take a truly transformative product to replace them. Voicebuds won't support Apple's "Hey, Siri" command, since that involves having one of Apple's proprietary chips, though Chen says the company is integrating its own AI assistant. But if the Voicebud's transcription is as great as the company claims, they may eventually earn a coveted place in my pocket alongside Apple's buds.