TGSpeechBox

Category
Utilities

Description of App

A formant-based speech engine great for reading text at high speech rates in 26 languages. Built by a community of testers who improve each release.

TGSpeechBox is a high-quality speech synthesizer that works as a system voice with VoiceOver and as a standalone text-to-speech app.
Features:
• 26 languages with natural-sounding speech
• Full VoiceOver integration — use as your primary system voice
• 7 voice presets plus YAML-defined voice profiles
• Deep engine tuning: pitch mode, voice tone, formant sharpness, inflection, tremor, breathiness, and more
• Phoneme editor: fine-tune individual speech sounds with live preview
• Pronunciation dictionary: add custom word pronunciations with IPA override support
• Stress, compound, and character dictionaries for advanced control
• Export and share your customizations
• Adjustable sample rate (11025–44100 Hz)
• Rate override and rate boost for VoiceOver
• Language lock: keep your preferred language regardless of system setting
• Runs entirely on-device — no internet required, no data leaves your phone
TGSpeechBox uses a Klatt-style formant synthesis engine with eSpeak-ng for phoneme generation. It is open source under the MIT license.
Built by a blind developer, designed for accessibility from day one.

Version

N/A

Free or Paid

Free

Device(s) App Was Tested On

iPhone

iOS Version

26.4

Accessibility Comments

Fully accessible, but it would be great if we had sliders in addition to text fields.

VoiceOver Performance

VoiceOver reads all page elements.

Button Labeling

All buttons are clearly labeled.

Usability

The app is fully accessible with VoiceOver and is easy to navigate and use.

Other Comments

Here are some of the issues I am thinking of reporting to the developer:
  • The iOS app does not respect the pitch mode selection.
  • Some languages like Turkish and Russian lag behind eSpeak and don't have some of the later improvements, affecting intelligibility especially at high speech rates. Pronunciations for some phonemes in certain languages also degrade naturalness and intelligibility and make the voices even more robotic and harder to understand compared to eSpeak, diminishing any advantages TGSpeechBox offers. We should be able to customize and fine-tune all the phonemes and formants, and we should also be able to use all the eSpeak voices, including those that are somewhat more natural. The current klatt voices sound more like Eloquence, but lack one key feature that makes Eloquence the favorite choice for many blind users for decades: It remains highly intelligible, fluent, responsive and natural at high speeds. These Klatt voices, however, pronounce certain letters like "m" and "n" in a way that requires the user/listener to infer them from the content and context.
  • Trying to add a pronunciation dictionary entry for a non-English language results in the entry being added for English and the inability to delete it afterwards. I therefore had to uninstall and reinstall the app entirely.
  • Some of the parameters are quite confusing. Some I've been experimenting with, but some still remain a mystery to me.
  • The high-pitched voices sound weird.

App Store Link

https://apps.apple.com/us/app/tgspeechbox/id6759512621

Developer's Website

https://github.com/tgeczy/TGSpeechBox/

Recommendations

2 people have recommended this app

Most recently recommended by Enes Deniz 1 day ago

Options

Comments

By Brian on Sunday, April 5, 2026 - 12:29

I'm guessing this is an alternative and/or replacement for Eloquence and ESpeak. My question is, how well does it work on older devices?

TIA.

By Enes Deniz on Sunday, April 5, 2026 - 12:35

It does work fine on my iPhone SE 2022. Both eSpeak and TGSpeechBox should work fine on your device as well. These are lightweight and responsive TTS engines, and even Piper TTS works on my iPhone SE 2022.

By Brian on Sunday, April 5, 2026 - 13:21

I too have the iPhone SE 2022. Glad to hear this will work for me. 🙂

By Rocker on Sunday, April 5, 2026 - 14:23

Given the issues regarding the negative comments above I'd hold off for now!

By Enes Deniz on Sunday, April 5, 2026 - 15:32

Adjusting the parameters is easier on iOS compared to NVDA on Windows so I wanted to import the yaml file into NVDA but got no speech after importing the file and restarting NVDA. I had to uninstall and reinstall the add-on.