Neuphonic

Spec sheet

Product

Neuphonic builds voice AI for natural speech and voice interactions through on-device, on-premises, and cloud API deployments.Read 12 Oct 2026 · neuphonic.com

Intended users

The on-premises page identifies enterprises, financial institutions, healthcare providers, and government as ideal use cases.Read 12 Oct 2026 · neuphonic.com

Platforms

web,api,self_hostedRead 20 Sep 2026 · neuphonic.com

Free plan

YesRead 20 Sep 2026 · neuphonic.com

Integrations

The developer documentation lists integrations for Twilio, LiveKit, Pipecat and VideoSDK.Read 12 Oct 2026 · docs.neuphonic.com

Security

The privacy policy says safeguards may include access controls, encryption, monitoring and secure infrastructure.Read 12 Oct 2026 · neuphonic.com

Privacy

The maker says on-device processing keeps voice data local and on-prem deployment gives organizations control over sensitive data and infrastructure.Read 12 Oct 2026 · neuphonic.com

Support

The maker invites customers to contact its team to discuss integrations and on-device or on-premises deployment.Read 12 Oct 2026 · neuphonic.com

Headquarters

London, United KingdomRead 28 Sep 2026 · neuphonic.com

API access

YesRead 20 Sep 2026 · neuphonic.com

Audio codec

NeuCodec encodes and reconstructs audio, and the API page describes a speech codec with near-inaudible reconstruction loss.Read 12 Oct 2026 · neuphonic.com

Commercial licensing

Commercial terms vary by model; NeuDecide and NeuCodec use Apache 2.0, while NeuTTS-2E uses the NeuTTS Open License with commercial-use conditions.Read 12 Oct 2026 · neuphonic.com

Commercial use

YesRead 20 Sep 2026 · neuphonic.com

Company identity

The privacy policy identifies the company as Neuphonic Limited and lists its address as Office 3 Cottam House, 36-40 York Way, London, N1 9AB, United Kingdom.Read 12 Oct 2026 · neuphonic.com

Developer access

The maker directs developers to its API playground, downloadable models on Hugging Face, and code on GitHub.Read 12 Oct 2026 · neuphonic.com

Developer tools

Neuphonic offers Python and JavaScript SDKs; its quickstart specifies Python 3.9+ and Node.js 18+.Read 12 Oct 2026 · docs.neuphonic.com

Device support

The on-device solution targets mobile, desktop, embedded devices and wearables, and supports integration by SDK, API or custom deployment.Read 12 Oct 2026 · neuphonic.com

Export formats

WAVRead 20 Sep 2026 · neuphonic.com

Integration

The on-device solution can be integrated through an SDK, API, or custom deployment for mobile, desktop, embedded, and wearable hardware.Read 12 Oct 2026 · neuphonic.com

Licensing

Commercial terms vary by model; NeuDecide and NeuCodec use Apache 2.0, while NeuTTS-2E uses the NeuTTS Open License with commercial-use conditions.Read 12 Oct 2026 · neuphonic.com

Model limits

NeuTTS-2E provides English speech with seven emotion modes and four fixed speakers, and does not yet support voice cloning.Read 12 Oct 2026 · neuphonic.com

Model portfolio

The site describes NeuTTS-2E as English speech with seven emotion modes and four fixed speakers, and NeuDecide as converting speech into tool calls.Read 12 Oct 2026 · neuphonic.com

Offline use

On-device and on-prem deployments can run offline after installation, while the cloud API requires an internet connection.Read 12 Oct 2026 · neuphonic.com

On-device

On-device models run inference directly on user hardware and can operate offline after installation.Read 12 Oct 2026 · neuphonic.com

On-prem deployment

On-prem deployments run inside an organization’s own servers, private cloud or secure infrastructure, keeping data within that infrastructure.Read 12 Oct 2026 · neuphonic.com

On-premises

On-premises deployments run within an organisation’s own servers or private cloud, keeping data within its infrastructure.Read 12 Oct 2026 · neuphonic.com

Text to speech

Its API provides real-time text-to-speech with low latency, flexible voices, pronunciation controls, and support for multilingual speech.Read 12 Oct 2026 · neuphonic.com

Voice cloning

The cloud API offers voice cloning, and NeuTTS-2M offers voice cloning in seven listed languages for infrastructure deployments.Read 12 Oct 2026 · neuphonic.com

Where Specifiction places it

• Text-to-speech tools#121 of 2337.9
• Text-to-Speech Software#145 of 2237.7