GA Neuphonic

Neuphonic API

UK voice company focused on on-device and on-prem TTS (NeuTTS open models that run on CPU) with a hosted SSE/WebSocket API in 37 languages. API pricing is not published.

Est. per minuten/a

Overview

Best for: Teams that want the same voices hosted and on-device/on-prem (CPU) for privacy.

At a glance

CloningYes
Latency ms25
Languages37
WebRTCNo
WebSocketYes
gRPCNo
Self-hostYes
Open weightsYes

Sub-25 ms is a vendor figure cited by third parties; an independent benchmark measured ~763 ms. 37 languages per homepage, model spec lists 7. Pricing quote-based. Open weights are the NeuTTS Air/Nano models. Also SSE.

Audio in

Text with inline <pause> and <phoneme> tags

Audio out

Not verified

Languages

37 (homepage playground); NeuTTS-2M spec lists 7

Voices

Instant voice cloning

Latency

Third-party listings cite a vendor sub-25 ms figure; Vapi Humanness Index measured ~763 ms for the model it tested.

Regions

Not specified

Compliance

Not verified.

Features

  • SSE and WebSocket streaming
  • voice cloning
  • phoneme-level pronunciation tags
  • on-device and on-prem deployment

Pricing

WhatPriceUnitNotes
APINot publishedContact sales per third-party listings
How the per-minute estimate was worked out

Pricing not published

Free tier: Playground available; API free tier not verified

Source: neuphonic.com

Setup

  1. Get an API key from app.neuphonic.com.
  2. pip install pyneuphonic (Python) or @neuphonic/neuphonic-js.
  3. Use the SSE client for one-shot streaming or the WebSocket client for multi-chunk input.

Endpoint

See docs.neuphonic.com

Authentication

API key

Warnings

No public price list

API pricing is quote-based; budget cannot be estimated from public pages.

Latency claims vary wildly

A vendor sub-25 ms figure circulates while an independent benchmark measured ~763 ms; test before choosing for live agents.

Language counts disagree

Homepage says 37 languages; model spec lists 7 and directories list 7-10.

Plus 12 warnings that apply to all text-to-speech, streaming APIs. See category warnings.

Limits

  • Cloning samples: MP3/WAV, min 6 s, under 10 MB (third-party)

Models and products

NameStatusNotes
NeuTTS-2M / NeuTTS-2EListed on site~60M active params, 2x real-time on 2 CPU threads (site spec table).
neutts-air / neutts-nano (open weights)Open weightsSee the NeuTTS open-model entry.

Docs and sources

Docs

Sources used

Not fully verified

Pricing, output formats, exact endpoints, latency.

Similar text-to-speech APIs

Spotted a wrong price or a dead link?