praat/praat.github.io

Health v2 · maintenance only

Flags: no_license

Adoption not part of the score

1985 stars · 293 forks observed · 2026-09-25

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Praat is a desktop application for analyzing, synthesizing, and manipulating speech, widely used in phonetics research and teaching. It provides spectrograms, pitch/formant/intensity analysis, speech labeling with IPA, text-to-speech, and statistical and neural learning tools.

Use cases

• analyze pitch and formants in speech recordings
• generate spectrograms of audio files
• annotate and segment speech with IPA labels
• synthesize speech from pitch and formant curves
• manipulate pitch and duration of speech utterances
• run listening experiments for phonetics research
• measure voice quality jitter and shimmer

When to choose

• you need a free, mature tool for phonetic speech analysis
• you want publication-quality acoustic figures like spectrograms
• you need manual speech annotation with multi-tier labeling
• you teach or research linguistics and phonetics

When to avoid

• you need programmatic speech processing in a Python/ML pipeline rather than a GUI
• you need real-time or streaming audio analysis
• you want a modern collaborative or web-based tool

Facets

application · maturity stable

audio-processing speech-recognition tts data-visualization machine-learning data-science speech-processing education windows cross-platform phonetics spectrogram speech-analysis speech-synthesis annotation acoustics linguistics audio natural-language-processing research macos linux desktop

Member repositories

For agents

markdown · JSON · MCP: product_card(name="praat/praat.github.io")