Samplab
TextToSample is a free tool developed by Samplab that utilizes generative AI to convert text into audio samples. This tool allows users to ...
Last verified:
What is Samplab?
TextToSample is a free AI-powered plugin by Samplab that generates audio samples from text prompts using generative AI. Based on Meta's open-source MusicGen model, the tool lets musicians and producers type descriptions like
Samplab pricing
Pricing model: Freemium
TextToSample is completely free to download and use for non-commercial purposes under the CC BY-NC 4.0 license. The free tier includes: Samplab Plugin & Desktop App, up to 10 seconds per audio file, mono audio output, and real-time warping. Samplab Premium costs $12/month or $120/year (also listed as EUR 7/month or EUR 70/year), and includes: audio files of any length (fair use), stereo audio output, premium note controls (volume and pan), real-time and premium warping, and continuous updates. A 14-day free trial is available for Premium.
Samplab pros
- Completely free to use for non-commercial purposes
- Runs locally on your computer without internet connection
- Available as standalone app and VST3/AU plugin
- Generates samples instantly from text prompts
- Based on Meta's state-of-the-art MusicGen AI model
- Polyphonic note editing for audio like MIDI
- Chord detection and automatic pitch shifting
- Stem separation into drums, bass, vocals, instrumental
- Audio-to-MIDI conversion for polyphonic audio
- Can extend existing audio files with new prompts
- Automatic tempo and key adjustment for samples
- DAW integration with sync tempo and key linking
- Drum editing alongside harmonic notes
- Slice notes and add offsets precisely
- Stretch notes to any length without quality loss
- Multi-track editing capability
- Export cleaned-up MIDI files
- Works with Windows 10 and macOS 12+
- GPU or M1/M2 Neural Engine recommended for best performance
- No expiry date on the free tool
Samplab cons
- Non-commercial license only (CC BY-NC 4.0), no commercial use
- Free version limited to 10 seconds per audio file
- Free version outputs mono audio only, not stereo
- Requires powerful hardware (GPU or M1/M2 recommended)
- Only supports VST3, no VST2 version available
- No AU format for Windows users
- Computationally intensive processing on local machine
- Only supports Windows 10 and macOS 12 or higher
- No mobile application available
- Model weights use only the 'small' MusicGen model
- Training data not publicly disclosed
- No collaborative work features
- Cloud-based stem separation may require internet initially
- Limited to 64-bit DAWs only
- No built-in sample library or preset sounds
Frequently asked questions about Samplab
Do I have to pay to use TextToSample?
No, you do not have to pay. TextToSample is completely free to use. The developers released it as a free plugin to allow the producer community to experiment with state-of-the-art generative AI models, particularly showing what's currently possible with these models when running locally on your computer.
What is the purpose of TextToSample?
The purpose of TextToSample is to enable the producer community to experiment with state-of-the-art generative AI models. The developers want to show what is currently possible with these models, especially when running them locally on your computer. TextToSample is not intended to replace human musicians, but rather to provide a tool for experimentation and creative sample generation.
How is TextToSample licensed?
TextToSample uses Meta's MusicGen model weights, which are licensed under CC BY-NC 4.0. Specifically, the weights of the 'small' model are used. This license means the tool is only available for non-commercial purposes. You must give attribution to the creator, and only noncommercial uses of the work are permitted.
What data was this trained with?
TextToSample uses Meta's MusicGen model weights, specifically the 'small' model version. The MusicGen model is Meta's state-of-the-art open-source AI music generator. While the specific training data details are not fully disclosed in available documentation, MusicGen is an open-source model developed by Meta for AI music generation.
Do I need an Internet connection?
No, you do not need an internet connection to use TextToSample. All processing runs on your own computer locally. You only need internet access to download the models when you open TextToSample for the first time. After the initial download, the tool works completely offline.
What hardware do I need?
TextToSample has intensive computational and processing requirements. The developers expressly recommend using latest-generation chips such as Intel Core i9 13900K or i7 13700K, as well as Apple M1 with Neural Engine. A modern graphics card or GPU also helps significantly. All processing is done on your own machine, so it works best if you have a GPU or an M1/M2 Mac with Neural Engine.
Which operating systems do you support?
TextToSample currently supports Windows 10, macOS 12, and higher versions of both operating systems. The plugin is available in VST3 and AU formats, compatible with 64-bit DAWs on both Windows and Mac systems. It is not available for older operating systems or 32-bit systems.
Can I use TextToSample for commercial music production?
No, you cannot use TextToSample for commercial purposes. The tool is licensed under CC BY-NC 4.0, which explicitly permits only non-commercial uses. This means you can use it for personal projects, experimentation, and learning, but not for any music or audio that will be sold, distributed commercially, or used in commercial products.
How do I generate samples with TextToSample?
Simply type what you'd like to hear as a text prompt - anything from 'anthemic lead synth line' to 'atonal harmonica solo' will work. TextToSample will generate a sample based on your instructions. You can also drag and drop any existing sample into TextToSample, add a prompt, and the software will create a new sample based on the original. The interface is clear and user-friendly, letting you type text corresponding to the sound you want to generate.
Is there a VST2 version of TextToSample?
No, there is no VST2 version available. TextToSample is available only as a VST3 plugin (and AU for Mac), plus as a standalone application. It does not support the older VST2 format, so you'll need a DAW that supports VST3 or AU plugins to use it within your audio production software.