Nirva Raises $33M for $249 Voice Isolating AI Pendant
Nirva raises $33 million for a $249 AI pendant that isolates its wearer's voice, deletes audio by default, and ships before the end of 2026.
Summary
Nirva announced a $25 million pre Series A round on October 22, 2026, following $8 million raised earlier from investors including Lightspeed and South Park Commons, for $33 million total. Founded in 2025 by former Meta Reality Labs employees Wei Lyu and Jason Chen and fashion jewelry founder Hanying Hu, the startup has built an eight gram pendant that doubles as a bracelet and lasts up to three days per charge. The $249 device will open for orders in November in silver, brushed silver, black and gold, with early deliveries targeted by year end and additional devices planned.
Nirva uses onboarding voice samples to identify the wearer, passively analyze only their speech and filter surrounding speakers. Before deleting audio and transcripts by default, it measures words, tone, pitch, volume and speed. Its app infers emotions, personality patterns, communication style, companions, locations and daily notes, then provides reminders and suggested actions, positioning Nirva around self understanding rather than productivity. Privacy concerns associated with always listening devices and Ray-Ban Meta remain, while competition includes Taya, Vocci, Sandbar Ring, Friend and Bee. The wider ambient AI field also includes Amazon's wrist worn Bee device, former Plaud executive led Ambi, which plans hardware alongside its apps, and Apple's Voice Intelligence features introduced at its September iPhone launch.
Positives
- $25 million in new funding lifts Nirva's total to $33 million, following earlier backing from Lightspeed and South Park Commons.
- Eight gram jewelry styling, bracelet compatibility and up to three days of battery life support continuous wear.
- Onboarding voice modeling is designed to exclude nearby speakers rather than capture entire conversations.
- Audio and transcripts are deleted by default after Nirva extracts speech and vocal signals.
- November orders and year end 2026 deliveries give Nirva a defined route to its first users.
Risks & concerns
- Always on passive listening could sustain privacy concerns despite Nirva's speaker filtering and default deletion policies.
- Emotion, personality and communication insights depend on inferences whose accuracy has not been detailed.
- Early users must wait until the end of 2026, and longer term device plans remain unspecified.
- Taya, Vocci, Sandbar Ring, Friend, Bee, Ambi, Amazon and Apple are pursuing overlapping ambient AI or wearable experiences.