ModelTalker

ModelTalker creates personal synthetic voices for people who are losing or have lost the ability to speak. Users record speech, listen to preliminary voice versions, then choose whether to pay for and download a completed voice. Recording is available through a web tool or the MTVR recorder, which runs only on Windows. The Gen3 inventory uses about 300 sentences; a standard inventory has over 3,000, and custom recording is also available.

Voices can be used on Windows, Mac, iOS, and Android devices. The free iOS app can install a voice as an iOS system voice, and a Mac app lets users add their personal voice alongside the default child voice Jamie. Windows voices conform to Microsoft SAPI 5 and work with compatible speech-aware software, but not Microsoft Store apps using the other TTS standard. ModelTalker lists Predictable and ChatAble from TherapyBox and TD Talk from Tobii Dynavox as compatible iOS apps. The one-time fee is $100 before download; donors and clinicians can create voices without it. Copies are currently available for up to 18 months.

ModelTalker offers a recording-based path to a personal synthetic voice, with several device options and voice inventories. Check the recording method, supported software, and copy availability period before choosing it.

ModelTalker suits people who want to preserve a personal synthetic voice for communication, along with clinicians supporting that process. It may also suit users who need a voice for supported Windows, Mac, iOS, or Android devices.

ModelTalker builds a synthetic voice from recordings for people who are losing or have lost the ability to speak. It suits people who want a personal voice for supported devices and communication software. Its recording and device options are useful, but the Windows-only recorder and 18-month copy availability period deserve close attention.

Overview

Voice banking starts with recording speech, then lets users audition preliminary versions and decide whether to pay for and download a completed voice. The Gen3 inventory uses about 300 sentences; a standard inventory of over 3,000 sentences and custom recording are also available. The shorter inventory offers a more compact recording route, while the larger one and custom option give users different ways to build a voice.

The $100 fee covers help getting started, answers to questions, voice construction, and an archived backup. Recordings made through the online tool are stored securely on ModelTalker servers, and personal information is not shared outside the project without written permission. Completed voice copies remain available for up to 18 months, a meaningful limit for anyone who may need to download or reinstall a voice later.

Key features

Users can record with either the Windows-only MTVR recorder or a web-based recording tool. The web option is the practical choice for people without a Windows computer; those using MTVR need access to Windows for recording.

Windows voices conform to Microsoft SAPI 5, so compatible speech-aware software can use them. Microsoft Store apps that use Microsoft's other text-to-speech standard are not supported. On iOS, ModelTalker voices can be installed as system voices, and the maker lists Predictable, ChatAble from TherapyBox, and TD Talk from Tobii Dynavox as compatible apps.

ModelTalker also offers an app for Mac laptops and desktops, where users can add a personal voice alongside the default child voice Jamie. Voices are available for Windows, Mac, iOS, and Android devices. The product supports one language, so it is a poor fit for someone who needs a voice in multiple languages.

Pricing

The Personal voice plan is $100 a month, billed as a one-time service fee payable before downloading the completed voice. It includes voice construction and assistance, auditioning completed voice versions, and a download. Despite the plan's monthly label, the charge is one-time, not a recurring subscription.

Personal voice banking is $100, paid once, also a one-time fee payable before downloading the completed voice. Registration, recording, and auditioning are free; the fee covers downloads for different operating systems and devices, with copies available for up to 18 months. This is the more suitable option for someone who wants to complete the recording and audition process before committing payment.

Donors and clinicians can create voices without the $100 fee. Donors may download one copy for personal use. ModelTalker also offers free access through its iOS app, which can install voices as system voices; the completed voice download fee still applies where relevant.

Platforms

ModelTalker supports Android, iOS, Mac, web, and Windows. Recording is available through the web tool or the Windows-only MTVR recorder. Windows voices work with SAPI 5-compatible speech-aware software, while the iOS app installs voices as system voices. The Mac app supports adding a personal voice alongside Jamie.

Who it's for

ModelTalker is aimed at people who are losing or have lost speech and want to preserve a personal synthetic voice. It is strongest for users who can record with the web tool or Windows recorder and whose chosen device and communication software fit its supported options. It is less suitable for users who need multiple languages, Microsoft Store TTS apps, or voice copies beyond the 18-month availability period.

Pros and cons

• Pros: Users can audition preliminary voices before paying for a completed voice, making the fee a later decision.
• Pros: Web recording avoids the Windows-only MTVR requirement, and voices support Windows, Mac, iOS, and Android devices.
• Pros: The fee includes assistance, voice construction, and an archived backup, rather than only a download.
• Cons: A single supported language limits its usefulness for multilingual communication.
• Cons: Windows compatibility excludes Microsoft Store apps that use Microsoft's other TTS standard.
• Cons: Archived copies are available for up to 18 months, so users need to plan downloads and device setup within that period.

Alternatives

For broader voice-cloning and text-to-speech options, browse Voice Cloning Software and Text-to-Speech Software.

Voicebox is a freemium option with a free plan and support for Android, iOS, Linux, macOS, Windows, API, and self-hosted use. CereProc is a paid service with no free plan and a broader platform range, including web, Linux, API, and self-hosted use. Fish Audio has a free plan and supports API, web, Linux, macOS, Windows, and self-hosted use. Kits AI has a free plan and supports API, web, and Windows.

Qwen3-TTS is a free open-source option with 0.6B and 1.7B models and support for 10 languages. Sprag is a paid option for stock-voice text-to-speech, with Chorus pricing at $0.70 a month billed at $0.70 per hour of speech generated. VoiceStudio is a free option. Chatterbox TTS offers open-source models for research and development under the MIT license, with support for Linux, macOS, and self-hosted use.

Verdict

Choose ModelTalker if you need to preserve a personal voice for someone losing speech and its recording routes, devices, and compatible communication software suit your setup. Its audition-before-payment process and included assistance support that focused use. Look elsewhere if multilingual voices, Microsoft Store TTS compatibility, or long-term access to archived voice copies are essential.