Neuphonic was founded by Jiameng Gao, a former co-founder of dubbing company Papercup, and Sohaib Ahmad, a former hedge fund quant trader. The two met at the University of Cambridge while studying machine learning, and built Neuphonic around a focus that many voice AI companies treat as secondary: raw speed. In conversational voice, latency is the difference between an interaction that feels human and one that feels like talking to a slow machine, and Neuphonic's core innovation is a patent-pending approach to real-time, incremental speech generation.
Rather than synthesizing a full utterance before playback, Neuphonic's models generate speech incrementally, streaming audio with extremely low latency that the company has reported at around 25 milliseconds. This makes the technology well suited to live conversational agents, where every fraction of a second of delay degrades the experience. The company describes its text-to-speech as among the fastest in the world.
Neuphonic is also investing in multilingual coverage and in on-device voice solutions. On-device synthesis is strategically important because it removes dependence on cloud connectivity and lowers per-request cost, while improving privacy — a meaningful advantage for embedded, wearable, automotive, and offline use cases.
The company raised roughly €3.5M (about £3M) in pre-seed funding led by Moonfire Ventures, with participation from Tiny VC, the Salica Oryx Fund, and Cur8 Capital. Neuphonic said it would use the capital to expand language and voice coverage, deepen its research to improve model performance, and advance its on-device offering. With a strong technical pedigree and a clear differentiator in latency, Neuphonic is positioning itself as a voice infrastructure layer for the next wave of real-time conversational products.