FlowState can generate custom sounds using AI. Describe what you want, and AI creates unique audio—melodies, sound effects, textures, and more. No sample hunting, no synthesis expertise needed.
What Can AI Generate?
- Melodies — Original melodic phrases in any style
- Sound Effects — Risers, impacts, transitions, textures
- Ambient Sounds — Pads, drones, atmospheres
- Musical Phrases — Chord progressions, arpeggios
- Textures — Noise, vinyl crackle, background elements
How to Generate Sounds
Using Voice Commands
-
Click the microphone button
Find it in the transport bar. -
Say what you want
"Generate a dark synth melody in C minor" -
Wait for generation
AI creates the sound (usually 10-30 seconds). -
Review and use
The sound appears on your timeline. Keep it or regenerate.
Using the AI Panel
-
Open AI Studio
Click the AI button or use the command palette (Cmd/Ctrl+K). -
Choose generation type
Select "Generate Sound" or "Generate SFX". -
Describe your sound
Type a detailed description. -
Set parameters
Choose duration, key (for melodic content), and style. -
Generate
Click Generate and wait for results.
Writing Good Prompts
Better descriptions = better results. Include:
The Basics
- What: Type of sound (melody, pad, riser, etc.)
- Style: Genre or reference (trap, lo-fi, cinematic)
- Mood: Emotional quality (dark, uplifting, melancholic)
Optional Details
- Key: Musical key (C minor, F major)
- Tempo context: Fast, slow, mid-tempo
- Instrument: Piano, synth, strings, etc.
- Texture: Clean, distorted, lo-fi, crispy
Example Prompts
Melodies
dark trap melody in D minor, simple and catchy
lo-fi piano loop, jazzy chords, relaxed feel
uplifting synth lead, major key, energetic
melancholic guitar melody, slow and emotional
Sound Effects
big riser for a drop, builds tension over 8 bars
impact hit for a beat drop, cinematic and punchy
transition whoosh, quick sweep left to right
glitchy transition effect
Textures & Ambience
vinyl crackle and noise, subtle background texture
dark ambient drone, ominous and evolving
warm analog pad, sustained and atmospheric
rain and thunder ambience
Generation Settings
Duration
- Short (1-4 bars) — One-shots, short phrases
- Medium (4-8 bars) — Loops, standard sections
- Long (8-16 bars) — Extended phrases, ambience
Musical Key
For melodic content, specify a key to match your project:
- If your song is in C minor, generate in C minor
- Relative keys work too (C minor / Eb major)
- Let AI choose if you're not sure—you can pitch shift later
Tip: If the generated sound isn't perfect, don't give up. Regenerate with a slightly different prompt, or use the sound as a starting point and edit it.
Working with Generated Sounds
After generation, you can:
- Keep it — It's already on your timeline
- Regenerate — Try again with the same or different prompt
- Edit — Trim, chop, or process like any other clip
- Layer — Combine with other sounds
- Pitch shift — Change the key to fit your song
Best Practices
- Be specific — Vague prompts give vague results
- Include musical context — Key, tempo, style
- Iterate — Generate multiple options and pick the best
- Combine with editing — AI gives you raw material to shape
- Use as inspiration — Let AI spark ideas you can develop
Generation Limits
FlowState includes generous AI generation:
- Certain number of generations per day (depends on your plan)
- Each generation takes 10-30 seconds
- Generated sounds are yours to use royalty-free
Troubleshooting
Generation Failed
- Check your internet connection
- Try a simpler prompt
- Wait a moment and try again
Result Doesn't Match Prompt
- Be more specific in your description
- Try different wording
- Generate multiple options