Generate ambient music from text. Locally. No GPU required. - Read more about how it works.
latentscore_2026-03-01.mp4
Four ways:
- π§ Hear it now - latentscore.com/demo. Browser, no install.
- π³ Run the demo + JupyterLab locally -
docker composebuilds the full demo from source in an isolated environment β demo UI and a JupyterLab playground, no local Python needed. ~15 min first build. See Try the demo. - π§ Open in Colab - free CPU runtime, no local setup, no key required.
- π Build with it locally -
pip install latentscoreon macOS or Linux (Python 3.11/3.12). Secondary path β use Docker for the full demo. See Try the SDK.
import latentscore as ls
ls.render("morning coffee shop").play()That's it. One line. You get audio playing on your speakers.
No install needed: open latentscore.com/demo in your browser.
Run it locally β tested on Linux, macOS, and Windows with Docker Desktop (WSL2 backend on Windows).
Build from source (~15 min first time, near-instant after):
docker compose -f demo/docker-compose.yml up --build
β οΈ Apple Silicon (M-series) Macs: the build defaults tolinux/amd64for reproducibility, which runs under emulation and is slow. For a native, much faster build, prefix the command:LATENTSCORE_DOCKER_PLATFORM=linux/arm64 docker compose -f demo/docker-compose.yml up --build
The first prompt after build takes ~30 seconds as model weights load; subsequent responses are near-instant.
Then open:
localhost:4242- demo UIlocalhost:8889- JupyterLab playground (SDK pre-installed)
See the demo documentation for architecture, dev setup, and troubleshooting.
- Try it now - four ways to get going (browser, Docker, Colab, pip)
- Try the demo - deeper Docker setup
- Try the SDK - Colab (recommended) or local install
- CLI -
latentscore doctor,download,demo - Quick start - Python in 5 lines
- Controlling the sound -
MusicConfigparameters - Read more - docs, FAQ, research pipeline, contributing
- Citation - SIGGRAPH Talks '26 BibTeX
- Documentation - streaming, live playlists, async, bring-your-own-LLM
- FAQ - common questions
π³ For the full demo, use Docker. The
pip installpath is for library/scripting use on macOS and Linux.π§ Easiest SDK path: Open the Colab notebook β free CPU runtime, no install, no system deps to wrangle.
If you'd rather install locally, read on.
β οΈ Local pip installs may require system audio libraries depending on your OS. If you hit issues, use Docker or Colab instead β both include all dependencies.
- OS β macOS or Linux. Windows: use Docker Desktop with the WSL2 backend for the demo. Native Windows pip is unsupported, and WSL2 pip is not part of the supported SDK path.
- Python 3.11 or 3.12 - local SDK installs require this range via package metadata. For other Python versions, use Docker or Colab.
With venv (regular Python):
python3.12 -m venv .venv
source .venv/bin/activate
pip install latentscoreWith conda:
conda create -n latentscore python=3.12 -y
conda activate latentscore
pip install latentscoreNote
Linux: smaller CPU-only install. PyTorch's default Linux wheel bundles ~5 GB of CUDA libraries. LatentScore is CPU-only, so you can skip them with PyTorch's CPU index:
pip install latentscore --extra-index-url https://download.pytorch.org/whl/cpumacOS is CPU-only already, so plain pip install is fine there.
The default install gives you text prompts, full parameter control, and local playback. See the Quick start below.
Optional extras:
| Install | Adds |
|---|---|
pip install "latentscore[external]" |
bring-your-own hosted LLM via LiteLLM (Anthropic, Gemini, OpenAI, β¦) |
pip install "latentscore[heavy]" |
CLAP audio-based retrieval (fast_heavy model) |
π§ pip install "latentscore[expressive]" |
local LLM inference. Extremely experimental, under testing. |
latentscore doctor --strict --offlineExits non-zero with a clear hint if anything's broken. Add --json for machine-readable output.
β οΈ The CLI ships with the SDK install β same system-deps caveats as Try the SDK above. Skip the headache by running the Colab notebook instead; most of these CLI commands have equivalent Python calls (ls.prefetch(...),ls.render(...)) you can run there.
# Verify your install
latentscore doctor # human-readable summary
latentscore doctor --strict --offline # nonzero exit if anything's broken
latentscore doctor --json # machine-readable output
# Pre-download model assets (otherwise the first render call appears to hang)
latentscore download fast # ~90 MB, MiniLM embedding model
latentscore download fast_heavy # ~1.8 GB, LAION-CLAP weights
# Render a sample clip
latentscore demo # play a short ambient clip
latentscore demo --duration 30 --output ambient.wav # 30 seconds, save to fileπ§ Don't want to install? Try the same code in Colab β free CPU runtime, no install.
import latentscore as ls
# Optional one-time setup: pre-download the embedding model (~90 MB) so
# the first render() call doesn't appear to hang. The download happens
# on the first render anyway; this just makes it explicit and visible.
ls.prefetch("fast")
audio = ls.render("morning coffee shop", duration=10.0)
audio.play() # plays on your speakers
audio.save("output.wav") # save to WAVls.render("morning coffee shop").play()
ls.render("thunderstorm on a tin roof").play()
ls.render("tension over a treasured object").play()Beyond text prompts, you can drive synthesis directly:
import latentscore as ls
# Full control: build a MusicConfig with human-readable labels
config = ls.MusicConfig(
tempo="slow",
mode="dorian",
root="d",
bass="drone",
pad="ambient_drift",
melody="contemplative",
rhythm="minimal",
texture="shimmer",
echo="heavy",
density=3,
brightness="dark",
space="vast",
)
ls.render(config, duration=10.0).play()
# Or start from a vibe and slam knobs to opposites
ls.render(
"morning coffee shop",
update=ls.MusicConfigUpdate(
tempo="very_fast",
brightness="very_dark",
echo="infinite",
),
).play()See the full documentation for the parameter reference, relative-step updates, streaming, live playlists, async API, and bring-your-own-LLM cookbook.
- Documentation - parameter reference, streaming, async API, bring-your-own-LLM
- FAQ - first-call hang, system deps, Windows, citation, β¦
- Retrieval map card - what powers the default
fastmodel: row schema, construction pipeline, models used, biases, license - Research pipeline - how the dataset was built
- Contributing - setup + style rules
- Demo - run the web demo locally
If you use LatentScore in your research, please cite the SIGGRAPH Talks '26 paper:
@inproceedings{gupta2026latentscore,
author = {Gupta, Prabal},
title = {LatentScore: Sketching Soundscapes with LLM-Distilled Retrieval for Procedural Synthesis},
booktitle = {SIGGRAPH Talks '26},
year = {2026},
publisher = {ACM},
doi = {10.1145/3799818.3812120}
}LatentScore is released under the Apache License 2.0.