Audio sources and credits

Practice calls in Dispatcher Prep use AI-generated voices. Digit drills use recordings of human speakers. No exercise uses a real 911 call, and every name, address and number in our scripts is fictional.

Kokoro voices (practice calls)

Call scenarios are voiced with Kokoro-82M by hexgrad, an open-weight speech model released under the Apache License 2.0. We wrote the scripts. The voices are synthetic and do not depict real people.

Free Spoken Digit Dataset

FSDD by Zohar Jackson (Jakobovski) and contributors; selected voice: Jackson. Source repository. Licensed under Creative Commons Attribution-ShareAlike 4.0 International. The selected WAV recordings are unchanged. Where a playback sequence is an adaptation of this audio, that adaptation is also offered under CC BY-SA 4.0. FSDD README and license statement.

LibriSpeech / LibriVox narration

LibriSpeech corpus prepared by Vassil Panayotov, Daniel Povey, Guoguo Chen and Sanjeev Khudanpur, derived from LibriVox readings. Dataset and source. Licensed under Creative Commons Attribution 4.0 International. We converted selected FLAC files to mono MP3 at 64 kbps, without cutting them. The recordings and source transcripts keep their source licenses; our exercises and explanations are separate work.

LibriSpeech license notice.

Fonts

DM Sans and Manrope are used under the SIL Open Font License: DM Sans license, Manrope license.

Recording inventory

Download the full manifest with transcripts, sources, changes, durations and checksums. Licenses do not imply endorsement by speakers, dataset authors or model authors.