Neural Synthesis
Describe the sound in your head, let 2G generate it.
A tiny neural sampler.
2G is a sampler instrument inspired by early-2000s mobile phones. Describe the sound in your head, let Neural Synthesis generate it, then edit the sample and play it on your keyboard.
FREE
Specifications
Loop analysis
Compares timbre, volume, and vibrato when determining loop points. Searches for regions that minimize abrupt shifts and short repetitions, smoothing transitions with crossfade.
Presents up to three suitable candidates when found. Select one and sustain the sample by holding the key.
Root note analysis
Compares pitch measurements across different sections of the sample. When it detects a consistent pitch, it measures the fine-tuning discrepancy and corrects it on playback. If no prominent pitch is found, the sample is marked as Unpitched.
Shape your sound
Trim the start and end of the sample. Adjust note entry and tail dynamics with Attack and Release. Alter tone using Brightness, Detune, Noise, Quality, and Formant controls.
Small speaker simulation
Models the narrow frequency range, resonances, and diaphragm excursion of mobile device speakers. During prolonged high-velocity playback, a voice coil heating model naturally compresses the sound.
Four voices share the same physical model; playing chords pushes the speaker collectively. 16 kHz mono defines 2G’s distinct handheld character, bypassable with a single switch.
Plans
$0
- Instrument
- Full
- Own samples
- Unlimited
- Generations 20 monthly generation credits exclusive to the first 200 users. One-time 20 credits thereafter.
- 20 / mo
- Sound Engine
- Basic
$10/mo
- Instrument
- Full
- Own samples
- Unlimited
- Generations
- 200 / mo
- Sound Engine
- Studio
1 credit = 1 successful generation. Unused Studio credits do not roll over.
Frequently asked questions
What is Neural Synthesis?
The section that generates audio from your text descriptions using AI. The generated audio can then be edited and played as a sample inside 2G.
Can I use it without generating sounds?
Yes. You can import, edit, and play your own audio files. These operations consume zero credits and do not require a Studio subscription.
Which file formats can I import?
WAV, AIFF, FLAC, and MP3. Up to 20 MB in file size and 10 seconds in duration. Your original source file remains untouched.
Is an internet connection required?
Required only for Neural Synthesis and credit verification. Local samples can be played and edited completely offline.
What happens if I run out of credits?
Credits are only needed to generate new sounds. You can continue using your saved sounds and your own samples indefinitely.
What happens if I cancel my subscription?
Generation access continues until the end of your paid billing period. The local sampler and all saved sounds remain permanently accessible.
Can I use the generated sounds in my music?
Yes. Generated sounds are royalty-free for personal and commercial music. Just don’t repackage or resell them as standalone samples or libraries.
Installation requirements for 2G
Windows
Windows 10 or later (64-bit)
Any 64-bit multi-core processor
50 MB free disk space
VST3
STANDALONE
macOS
macOS 11 Big Sur or later
Apple Silicon & Intel compatible
50 MB free disk space
AU
VST3
STANDALONE
changelog · 1.0.4
1.0.4 · 2026-10-24 20:41 1.0.4 FULL LOOP: THE WHOLE SOUND LOOPS BACK TO THE START 1.0.4 PROMPT ENHANCE, ON / OFF IN OPTIONS 1.0.4 SYSTEM PROMPT SETTING REMOVED 1.0.4 NEW START-UP ANIMATION 1.0.3 · 2026-10-22 12:07 1.0.3 FREE LICENSE SETS ITSELF UP, NO CODE TO ENTER 1.0.3 EXPIRED STUDIO CODE FALLS BACK TO FREE 1.0.2 · 2026-10-22 00:12 1.0.2 NEWER SOUND GENERATION MODEL 1.0.2 PHONE SPEAKER OFF BY DEFAULT, B TOGGLES IT 1.0.2 SETTINGS GROUPED: SOUND / PHONE / ACCOUNT 1.0.1 · 2026-10-20 16:15 1.0.1 AUTO PROMPT MODE IS THE DEFAULT 1.0.1 CLEANER SINGLE-NOTE GENERATIONS 1.0.1 ROOT NOTE DETECTED AFTER EACH GENERATION 1.0.1 ATTACK KNOB 0.5 MS - 3 S 1.0.1 FIXED CHECK SETUP AFTER A SUCCESSFUL GENERATION 1.0.1 UPDATE CHECK IN SETTINGS
