S
Code-Based Audio Synthesis
Your words. Your samples.
Free AI-generated samples and loops for beatmakers. Describe a kick, a bass, a pad — get a royalty-free WAV in seconds. No packs. No licensing. Just your sound.
royalty-freereal-timeai-poweredunlimited44.1khz wavno samplesown everythingroyalty-freereal-timeai-poweredunlimited44.1khz wavno samplesown everything
How should I write my prompt?
Be specific. For one-shots (kicks, snares, stabs), say "one shot" and optionally a duration in seconds (10s max). For loops, specify BPM and length in bars — e.g. "techno kick, 130 bpm, 4 bars". You can request a musical key like Cm or F#m, though results aren't guaranteed. The more precise your prompt, the better the output.
Are the generated sounds royalty-free?
Yes. Every sound generated by SampleLoom is synthesized from scratch using mathematical functions — no samples, no pre-recorded audio, no third-party libraries. You own the output and can use it in any commercial or personal project without attribution, licensing fees, or restrictions.
What types of sounds can SampleLoom generate?
SampleLoom works best with electronic and synthetic sounds: kicks, snares, claps, hi-hats, basslines, pads, chords, brass stabs, acid lines, and experimental textures. You can generate one-shot samples or looped patterns with BPM and bar count. It is not designed for realistic acoustic instruments, vocals, or full song arrangements — it excels at drum hits, synth loops, and sound design elements.
How does the sound generation work?
You describe the sound you want in plain text. An AI model interprets your prompt and writes Python synthesis code that runs directly in your browser using WebAssembly. The audio is generated client-side at 44.1 kHz (CD quality) and output as a WAV file — nothing is processed on a remote server.
Can I refine a sound after generating it?
Yes. After each generation, SampleLoom suggests refinements like "darker", "more distortion", or "longer decay". Click any suggestion or type your own instruction and hit Refine. The AI keeps the original structure (BPM, duration, layers) and only modifies what you ask for, so you can iterate without starting over.
What format are the output files?
All sounds are exported as uncompressed WAV files at 44,100 Hz sample rate, ready to drag into any DAW (Ableton, FL Studio, Logic, etc.). No conversion needed.
Do I need to install anything?
No. SampleLoom runs entirely in your browser. The synthesis engine loads automatically on first visit (a few seconds for the WebAssembly runtime). No plugins, no downloads. A free account (just your email) is needed to refine sounds — no password required, we use magic link login.
What are the limitations?
SampleLoom generates sounds through additive and subtractive synthesis. It cannot load external samples, apply studio effects like reverb or delay, or produce realistic acoustic instruments. Very long sounds (over 30 seconds) may be slower to render depending on your browser. Results vary — sometimes you'll get exactly what you want on the first try, sometimes it takes a refinement or two.
Should I be careful with headphones or speakers?
Yes. AI-generated sounds can contain unexpected loud transients, heavy sub-bass, or harsh frequencies — especially kicks, bass, and experimental textures. Always start with your volume low and increase gradually. Be extra careful with headphones to protect your hearing.
Is SampleLoom free to use?
SampleLoom is currently free while in beta. Generate as many sounds as you want, download them, and use them however you like.
Can I drag and drop sounds into my DAW?
Yes. On desktop with Chrome, a "Drag to DAW" handle appears next to the Download button after each generation. Drag it directly into Ableton, FL Studio, Logic Pro, or any DAW that accepts WAV files. The file is automatically named after your prompt. This feature requires a mouse or trackpad — it is not available on touch devices.