Legal
Credits
Updated September 23, 2026
How these sources are used to make each set of words is described in How Orablo is made.
Pronunciations
Every IPA transcription in Orablo is derived from a published dictionary, or from spelling rules where the language allows it. Nothing is guessed.
The English transcriptions come from the CMU Pronouncing Dictionary.
Copyright © 1993–2015 Carnegie Mellon University. All rights reserved. Used under its BSD-style licence.
Spanish, Portuguese, Italian, French and German come from Wiktionary, read through the extracts published by Kaikki.org, and from the Montreal Forced Aligner pronunciation dictionaries. Spanish is also transcribed by Orablo’s own spelling rules, with the dictionaries as a check.
Wiktionary text © Wiktionary contributors, under the Creative Commons Attribution-ShareAlike 4.0 licence. Extracted with wiktextract by Tatu Ylonen, kaikki.org.
The pronunciation data Orablo adapted from Wiktionary via Kaikki.org is shared under the Creative Commons Attribution-ShareAlike 4.0 licence.
Montreal Forced Aligner dictionaries by Michael McAuliffe and contributors, under the Creative Commons Attribution 4.0 licence.
Listening on your device
When a browser has no speech recognition of its own, Orablo listens with Vosk, an offline speech recognition toolkit by Alpha Cephei, running on your device. The language models are Alpha Cephei’s small models, used under the Apache License 2.0, as is the vosk-browser library by Ciaran O’Reilly that runs them in the browser.
Word checks
WordNet is used while writing each English edition, to confirm a word exists, that the part of speech we claim for it is one it has, and that each definition renders one of its senses. The definitions themselves are Orablo’s own and are checked against those senses before they publish.
WordNet 3.0 Copyright © 2006 by Princeton University. All rights reserved.
Type
Headings are set in Fraunces by Undercase Type, everything else in Instrument Sans by Rodrigo Fuenzalida. Both under the SIL Open Font License.
Speech
Playback and listening use speech features built into your browser. What that means for your voice is in the privacy policy.