Skip to main content

kinescript-kokoro

Kokoro voices for kine voice. It speaks the text of a scene's sounds into WAV files and writes the time of each word, so layers can start on a word and captions follow the voice.

{"id": "vo1", "role": "voice", "text": "Write a scene. Check it. [[go]]Render it.", "speaker": "af_heart"}
kine voice scene.json     # writes voice/vo1.wav and voice/vo1.words.json
kine check scene.json

Install

Kokoro needs Python 3.10 to 3.12 and PyTorch. Into the same environment as Kinescript:

pip install kinescript-kokoro

From a checkout of Kinescript, pip install ./plugins/kinescript-kokoro does the same.

Kinescript itself runs on newer Pythons than Kokoro. If yours is one, install the plugin into another Python and point KINE_VOICE_CMD at it; kine voice sends it the lines as JSON on stdin and never through a shell:

python3.12 -m venv ~/kokoro-env
~/kokoro-env/bin/pip install kinescript-kokoro
export KINE_VOICE_CMD="$HOME/kokoro-env/bin/python -m kinescript_kokoro"

On Windows, set KINE_VOICE_CMD=C:\kokoro-env\Scripts\python.exe -m kinescript_kokoro. A JSON list such as ["C:/kokoro env/python.exe", "-m", "kinescript_kokoro"] also works and keeps a path with spaces whole.

The first run downloads the model (327 MB, plus about 0.5 MB per voice) from Hugging Face. On the development machine's CPU, a 4 s line took 9 s and 39 s of speech took 14 s, each including about 7 s to load the model.

Voices and speed

speaker is a Kokoro voice, af_heart unless the sound names one. Its first letter is the language: a American English, b British English (af_bella, am_adam, am_michael, bf_emma, bm_george, and the rest of the list in Kokoro's VOICES.md). rate is Kokoro's speed, 1 by default.

Word times come with English voices only. Other languages are spoken, but their sounds have no words, so word cues and voice captions do not work with them.

Device

It runs on the GPU when the installed PyTorch was built for it, and on the CPU otherwise (a GPU newer than the PyTorch build, such as an RTX 50 series card with a CUDA 12.4 build, would fail on the first line). KOKORO_DEVICE=cpu or KOKORO_DEVICE=cuda picks one.

Metadata

Release files for kinescript-kokoro 0.1.0

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

Source distribution (sdist)

Source distribution for kinescript-kokoro 0.1.0
File Size Uploaded Provenance Actions
kinescript_kokoro-0.1.0.tar.gz 5.6 kB With provenance Details for kinescript_kokoro-0.1.0.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for kinescript-kokoro 0.1.0
File Interpreter ABI Platform Actions
kinescript_kokoro-0.1.0-py3-none-any.whl Python 3 none any Details for kinescript_kokoro-0.1.0-py3-none-any.whl

Total release size: 11.9 kB

Release files / kinescript_kokoro-0.1.0.tar.gz

Download URL kinescript_kokoro-0.1.0.tar.gz
Size 5.6 kB
Tags Source
SHA-256 checksum
How to use checksums
c55942e1ef729b6855cb4ae57b0773dbfbfcd87febd4c216fdc71d30a07abd60
BLAKE2b-256 checksum
How to use checksums
b83c3e4236d68874e28382d9f3f5ab8b1a3e6393e4311bb2343cdc35b955f5eb
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 30, 2026.

Transparency log

Release files / kinescript_kokoro-0.1.0-py3-none-any.whl

Download URL kinescript_kokoro-0.1.0-py3-none-any.whl
Size 6.3 kB
Tags Python 3
SHA-256 checksum
How to use checksums
5ee9e4f08579b985ec14b95912a9ba3d64ab8e1054fc2bf30200c05e00913937
BLAKE2b-256 checksum
How to use checksums
6c0502fd05b59d2dd9836daebe2edc57ca5a439bc8d691b2e6f26a0c576ac352
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 30, 2026.

Transparency log

Security for kinescript-kokoro 0.1.0

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

Full provenance

Both files in this release have been uploaded with provenance metadata.

Select from 2 files

Attestation for kinescript_kokoro-0.1.0.tar.gz

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 30, 2026.

Transparency log

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Release history Release notifications | RSS feed

This release

0.1.0 This release

2 release files Full provenance

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page