Magic sound effects
Magic has no physical source, which makes it hard to prompt for. Describe a real material that can stand in for it.
Listen: 11 samples from the magic collection
Every clip shows the exact prompt that produced it. The render script reads its prompts from the same file as this page. Copy one, change a word and generate your own.
Arcane ritual chant
via audiocpp-ace-stepmystical fantasy ritual theme, ethereal chanting vocals, ancient choir, deep drone, arcane atmosphere
Lightning bolt spell
via audiocpp-stable-audio-3-sfxlightning bolt spell, sharp electric crack with a crackling arc tail and a rolling low rumble
Summoning portal
via audiocpp-stable-audio-3-sfxmagic portal opening, deep swirling drone rising with layered shimmering bells and a bright bloom
Wind gust spell
via audiocpp-stable-audio-3-sfxwind spell casting, sharp gust of air with a whistling upward sweep and a swirling tail
Poison spell
via audiocpp-stable-audio-3-sfxpoison spell, thick viscous bubbling with a sickly detuned drone underneath, close and wet
Barrier shatter
via audiocpp-stable-audio-3-sfxmagical barrier shattering, crystalline glass breaking into a bright shimmering scatter with a fading hum
Wand flick
via audiocpp-stable-audio-3-sfxmagic wand flick, short airy sparkle sweep with a bright twinkling tail, dry
Raise dead
via audiocpp-stable-audio-3-sfxnecromantic spell, low guttural drone rising with bone rattling and a whispered choral swell, cavernous
Mana recharge
via audiocpp-stable-audio-3-sfxmana recharging, soft pulsing energy shimmer building steadily with warm bell overtones, seamless loop
Transmutation
via audiocpp-stable-audio-3-sfxtransmutation spell, sparkling granular shimmer morphing into a solid metallic ring, close mic
Explosive rune
via audiocpp-stable-audio-3-sfxmagical rune detonating, bright energy snap into a deep resonant boom with a shimmering debris tail
Prompts that work
These models respond to physical description. Name the material, the action and the space.
Spell cast
magical spell cast, bright metallic shimmer, glass and bell overtones, rising pitch, sparkling texture, 1 second, close and dry
"Glass and bell overtones" gives the model two real resonant materials to imitate. The word "magical" has no acoustic signature of its own.
Charge-up
magical energy charging, low rumble building to a crackling hiss, fire and static layered, rising intensity over 1.5 seconds, close mic
Stack two source materials ("fire and static") to get a layered buildup. Either one alone sounds thin.
Ice burst
magical ice burst, glass shattering into fine crystalline shards, cold high-frequency sparkle, sharp attack, 800 ms, dry
"Glass shattering into fine crystalline shards" borrows a real physical event to suggest ice, which has no sound of its own.
Heal chime
magical healing chime, soft bell and glass resonance, warm sustained tone with gentle shimmer decay, 1.2 seconds, roomy
"Warm" and "gentle decay" set the mood. A healing spell needs a soft release envelope to sound benevolent; a hard cutoff sounds harsh.
Curse drone
malevolent magic drone, low detuned metal groan under a thin dissonant whine, no percussive attack, 2 seconds, roomy and unsettling
"No percussive attack" stops a dread drone from arriving as a hit. Dark magic should creep in.
Making magic sounds that fit
Describe the material magic is imitating
Magic has no acoustic signature of its own. Every convincing magic sound is a real material standing in for the effect: glass and bells for anything shimmering, fire and static for anything charging, metal groaning under strain for anything cursed. Pick the real material first, then describe what it is doing (shattering, resonating, straining). The word "magical" adds little.
Pitch direction tells the story
A rising pitch reads as gathering power, a falling pitch as release or dissipation, and a sudden cutoff as a spell being interrupted or landing. Say which one you want, such as "rising pitch" for a charge or "pitch cutting off sharply" for a teleport vanish. Otherwise the model cannot tell whether the spell is starting or ending.
Attach sparkle to a material
"Sparkling" and "shimmering" work as prompt words only when attached to a real high-frequency source: glass, bells, chimes, static. Used alone they tend to produce a generic bright hiss. Pair the texture word with its material every time, and apply it to one moment (the cast, the impact) instead of the whole sound.
Use ACE-Step for a chanted ritual
The sound-design samples above imitate a material standing in for magic: glass, fire, metal. A summoning ritual or an ancient incantation needs a chant with words. ACE-Step's `lyrics` field provides that ("[verse]\nAncient words\nAwaken now"). Use it when the spell is a ritual with words, and the SFX model for impact effects.
Why generate locally
The models run on your machine, which changes the cost, the privacy and what you can automate.
Offline after download
Weights are fetched once. After that, generation, the library and the API all work with the network off. Everything lives in one SQLite file and one folder of WAVs under ~/.soundbox, and nothing is uploaded.
Your agent can use it
Soundbox runs an MCP server in the same process, with tools for generating and listing sounds. Generated files come back as absolute paths, so an agent can copy them into a project.
Licences stated up front
Every model's terms are shown before you download it, and non-commercial weights need an explicit confirmation. Soundbox is MIT. The licence of the model that made a sound governs what you may do with it.
claude mcp add --transport http soundbox http://127.0.0.1:17494/mcpcurl -X POST http://127.0.0.1:17494/api/generate \
-H 'Content-Type: application/json' \
-d '{"model":"audiocpp-stable-audio-3-sfx","prompt":"8-bit coin pickup blip","duration_s":1}'Frequently asked questions
- Can I use these sounds in a commercial video or game?
- It depends on the model that made them. Soundbox shows each model's licence before you download its weights. The Stable Audio 3 models use the Stability AI Community License: commercial use is free below USD $1M annual revenue, and above that Stability requires an enterprise licence. ACE-Step 1.5 is MIT. MiniMax's models allow commercial use under a revenue cap and require an attribution notice. Some voice and stem-separation models are non-commercial only. Soundbox itself is MIT, but the audio is governed by the licence of the model that made it. Read that licence before you ship; this summary is not legal advice.
- Does this run offline?
- Yes, once the weights are downloaded. Soundbox runs the models locally and stores everything in one SQLite file and one folder of WAVs under ~/.soundbox. There is no account, API key or upload step. Only the initial model download needs the network.
- What hardware do I need?
- An NVIDIA GPU with 8 GB or more. CPU-only is supported, but each generation then takes minutes. That is fine for a few short clips and slow for a large batch.
- Can my coding agent generate these for me?
- Yes. While Soundbox is running it serves MCP at http://127.0.0.1:17494/mcp, with tools for generating and listing sounds. An agent can generate a sound as part of the task it is already doing and gets back an absolute file path to use in your project.
- How is this different from downloading a sample pack?
- A magic sample pack gives you someone else's idea of a fireball or a heal spell, usually tuned to a different game's art style. Generating lets you pick the real materials (glass, bells, fire, static) that match your spell's visual design, and regenerate when the art team changes the effect. It costs time and a GPU.