Leopard Binding for Python

Leopard Speech-to-Text Engine

Made in Vancouver, Canada by Picovoice

Leopard is an on-device speech-to-text engine. Leopard is:

Private; All voice processing runs locally.
Accurate
Compact and Computationally-Efficient
Cross-Platform:
- Linux (x86_64), macOS (x86_64, arm64), Windows (x86_64, arm64)
- Android and iOS
- Chrome, Safari, Firefox, and Edge
- Raspberry Pi (3, 4, 5)

Compatibility

Python 3.9+
Runs on Linux (x86_64), macOS (x86_64, arm64), Windows (x86_64, arm64), and Raspberry Pi (3, 4, 5).

Installation

pip3 install pvleopard

AccessKey

Leopard requires a valid Picovoice AccessKey at initialization. AccessKey acts as your credentials when using Leopard SDKs. You can get your AccessKey for free. Make sure to keep your AccessKey secret. Signup or Login to Picovoice Console to get your AccessKey.

Usage

Create an instance of the engine and transcribe an audio file:

import pvleopard

leopard = pvleopard.create(access_key='${ACCESS_KEY}')

transcript, words = leopard.process_file('${AUDIO_FILE_PATH}')
print(transcript)
for word in words:
    print(
      "{word=\"%s\" start_sec=%.2f end_sec=%.2f confidence=%.2f speaker_tag=%d}"
      % (word.word, word.start_sec, word.end_sec, word.confidence, word.speaker_tag))

Replace ${ACCESS_KEY} with yours obtained from Picovoice Console and ${AUDIO_FILE_PATH} to the path an audio file.

Finally, when done be sure to explicitly release the resources:

leopard.delete()

Language Model

The Leopard Python SDK comes preloaded with a default English language model (.pv file). Default models for other supported languages can be found in lib/common.

Create custom language models using the Picovoice Console. Here you can train language models with custom vocabulary and boost words in the existing vocabulary.

Pass in the .pv file via the model_path argument:

leopard = pvleopard.create(
    access_key='${ACCESS_KEY}',
    model_path='${MODEL_FILE_PATH}')

Word Metadata

Along with the transcript, Leopard returns metadata for each transcribed word. Available metadata items are:

Start Time: Indicates when the word started in the transcribed audio. Value is in seconds.
End Time: Indicates when the word ended in the transcribed audio. Value is in seconds.
Confidence: Leopard's confidence that the transcribed word is accurate. It is a number within [0, 1].
Speaker Tag: If speaker diarization is enabled on initialization, the speaker tag is a non-negative integer identifying unique speakers, with 0 reserved for unknown speakers. If speaker diarization is not enabled, the value will always be -1.

Demos

pvleoparddemo provides command-line utilities for processing audio using Leopard.

Metadata

Release files for pvleopard 3.0.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for pvleopard 3.0.1
File	Size	Uploaded
pvleopard-3.0.1.tar.gz	43.4 MB	Feb 13, 2026	Details

Built distribution (wheel)

Table of built distributions (wheels) for pvleopard 3.0.1
File	Interpreter	ABI	Platform	Reset
pvleopard-3.0.1-py3-none-any.whl 43.4 MB Feb 13, 2026	Python 3	none	any	Details

Total release size: 86.8 MB

Release files / pvleopard-3.0.1.tar.gz

Download URL	pvleopard-3.0.1.tar.gz
Size	43.4 MB
Tags	Source
SHA-256 checksum How to use checksums	`32d6187db2ad36ce170d14341c6e37251f29490b889ebaab65e2f860053cec0d`
BLAKE2b-256 checksum How to use checksums	`f5ce40fe395c0da6210761d8ae3a1768823224e6a6ffc9fc78f97875d7428216`
Upload date	Feb 13, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/6.2.0 CPython/3.12.3`

Release files / pvleopard-3.0.1-py3-none-any.whl

Download URL	pvleopard-3.0.1-py3-none-any.whl
Size	43.4 MB
Tags	Python 3
SHA-256 checksum How to use checksums	`afc3411f7128e631202626ab8a6db3afd0af207a858ab8b05daee2b3902de4b7`
BLAKE2b-256 checksum How to use checksums	`99a9df6956db0ac17a79254028f392106a92e34350905430f6f02ad2919265b8`
Upload date	Feb 13, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/6.2.0 CPython/3.12.3`