Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Magenta RealTime 2: An Open-Weights Live Music Model
| Date | Stars |
|---|---|
| 2026-08-02 | 1705 |
| 2026-08-06 | 1705 |
| 2026-08-12 | 1715 |
| 2026-08-18 | 1724 |
| 2026-08-19 | 1726 |
| 2026-08-20 | 1728 |
| 2026-08-21 | 1731 |
| 2026-08-23 | 1733 |
| 2026-08-25 | 1736 |
| 2026-08-26 | 1737 |
| 2026-08-27 | 1741 |
| 2026-08-28 | 1742 |
| 2026-08-30 | 1744 |
| 2026-08-31 | 1745 |
| 2026-09-01 | 1758 |
| 2026-09-02 | 1773 |
| 2026-09-03 | 1783 |
| 2026-09-04 | 1784 |
| 2026-09-05 | 1786 |
| 2026-09-06 | 1786 |
| 2026-09-07 | 1790 |
| 2026-09-08 | 1789 |
| 2026-09-09 | 1790 |
| 2026-09-10 | 1791 |
| 2026-09-12 | 1792 |
| 2026-09-13 | 1793 |
| 2026-09-14 | 1794 |
| 2026-09-15 | 1794 |
| 2026-09-16 | 1795 |
| 2026-09-17 | 1798 |
| 2026-09-18 | 1799 |
| 2026-09-20 | 1800 |
| 2026-09-21 | 1802 |
| 2026-09-22 | 1801 |
| 2026-09-23 | 1804 |
| 2026-09-24 | 1803 |
| 2026-09-25 | 1804 |
| 2026-09-26 | 1806 |
| 2026-09-27 | 1807 |
| 2026-09-28 | 1808 |
| 2026-09-29 | 1807 |
| 2026-09-30 | 1807 |
| 2026-10-02 | 1808 |
| 2026-10-03 | 1810 |
| 2026-10-04 | 1813 |
| 2026-10-05 | 1811 |
| 2026-10-07 | 1812 |
| 2026-10-08 | 1815 |
| 2026-10-09 | 1816 |
| 2026-10-10 | 1819 |
| 2026-10-11 | 1820 |
| 2026-10-12 | 1821 |
| 2026-10-13 | 1824 |
| 2026-10-14 | 1827 |
| 2026-10-15 | 1830 |
| 2026-10-16 | 1831 |
| 2026-10-17 | 1833 |
| 2026-10-18 | 1835 |
| 2026-10-19 | 1838 |
| 2026-10-20 | 1839 |
| 2026-10-22 | 1841 |
| 2026-10-23 | 1841 |
| 2026-10-24 | 1840 |
| 2026-10-25 | 1842 |
| 2026-10-26 | 1843 |
| 2026-10-27 | 1847 |
| 2026-10-28 | 1850 |
| 2026-10-29 | 1854 |
| 2026-10-30 | 1856 |
| 2026-11-01 | 1857 |
| 2026-11-03 | 1859 |
| 2026-11-04 | 1859 |
Today
— stars today
This week
+9 stars this week
This month
+48 stars this month
Momentum
0.0
growth rate 0.49%/day
# Magenta RealTime 2
[](https://github.com/magenta/magenta-realtime/actions/workflows/cli_tests.yml)
> [!NOTE]
> **Go [here](https://magenta.withgoogle.com/mrt2) for pre-built Apps & Plugins.**
Magenta RealTime 2 (MRT2) is a state-of-the-art open-weights model for real-time music generation. It contains several key components
* An [open-weights model](https://huggingface.co/google/magenta-realtime-2)
* A [Python library](README.md) `magenta-rt` for inference with JAX and MLX backends
* A [C++ inference engine](core/README.md) `magentart::core` for efficient streaming audio generation on Apple Silicon MacBooks
* A suite of [example applications](examples/README.md) built on the inference engine.
Use this project to run inference, build a DAW plugin, or embed the model into new applications of your imagination.
Future updates will support supervised fine-tuning.
📖 **Full documentation:** https://magenta.github.io/magenta-realtime/
(or build it locally — see [`docs/README.md`](docs/README.md)).
> [!NOTE]
> **Looking for Magenta RealTime v1?** The original model and code have been moved to the [`v1_legacy`](https://github.com/magenta/magenta-realtime/tree/v1_legacy) branch.
## Repo Highlights
- `magenta_rt/` — Python inference library (JAX / MLX backends).
- `core/` — C++ inference library (`magentart::core`).
- `examples/mrt2/auv3` — All-in-one AUv3 plugin for DAWs.
- `examples/mrt2/standalone` — All-in-one standalone macOS app.
- `examples/jam/` — App for exploring note control.
- `examples/collider/` — App for exploring prompt space.
- `notebooks/` - Notebook for trying Python API.
## Hardware requirements
**Real-time streaming** requires **Apple Silicon** (M-series). We offer two model sizes:
- **`mrt2_small`** (230M parameters) — runs real-time on any Apple Silicon Mac, including Air models.
- **`mrt2_base`** (2.4B parameters) — higher quality; requires a Pro Max chip for real-time streaming.
The table below shows which devices support **real-time streaming** (generating audio faster than playback):
| Device | `mrt2_small` (230M) | `mrt2_base` (2.4B) |
|---|---|---|
| M5 Max | ✅ | ✅ |
| M3 Max | ✅ | ✅ |
| M2 Max | ✅ | ✅ |
| M4 Pro | ✅ | ✅ |
| M2 Pro | ✅ | ❌ |
| M1 Pro | ✅ | ❌ |
| M4 Air | ✅ | ❌ |
| M3 Air | ✅ | ❌ |
| M1 Air | ✅ | ❌ |
> **Note:** Both models can also run **offline (non-real-time) inference** on any Apple Silicon Mac or NVIDIA GPU via the Python library. See more details on [`docs/models.md`](docs/models.md).
## Quickstart on Apple Silicon
```bash
# Install uv if you haven't and create a venv
curl -LsSf https://astral.sh/uv/install.sh | sh
uv venv --python 3.12
source .venv/bin/activate
# Install dependencies (Python dev)
uv pip install "magenta-rt[mlx]"
# Download resources: style model and codec model
# (i.e., MusicCoCa and SpectroStream)
mrt models init
# Download the streaming model you want to use
mrt models download
# Generate 4 seconds of music (change to `mrt2_small` for small model)
mrt mlx generate --prompt "disco funk" --duration 4.0 --model=mrt2_base
```
### Python Development
For local development, clone the repo instead of installing from PyPI:
```bash
git clone --recurse-submodules https://github.com/magenta/magenta-realtime.git
cd magenta-realtime
uv pip install -e ".[mlx]"
```
### C++ App Development
To get started building C++ apps, perform the following setup:
```bash
# Install dependencies
uv pip install "cmake<3.28"
# Build hello_mrt2 (a basic command-line interface)
cmake . -B build
cmake --build build --target hello_mrt2 -j10
# Generate 4 seconds of music
./build/examples/hello_mrt2/hello_mrt2 \
~/Documents/Magenta/magenta-rt-v2/models/mrt2_base/mrt2_base.mlxfn \
~/Documents/Magenta/magenta-rt-v2/resources \
100 \
--prompt "ambient pads with sub bass"
```
See the full documentation:
- [Installation](docs/installation.md)
- [Models & checkpoints](dExcerpt of 5,251 characters
Read on GitHub29
10
6
3
3
2
2
2
1
Dinghao Zhou
1
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:ec3433f414c61f29, topic:jax