Choose whether you need a one-time sound effect or a seamless ambience loop.
Describe the one-time sound effect in English. 4–2500 characters.
Output duration from 1 to 60 seconds. Token cost updates automatically.
Sign in to spend tokens
Describe the sound is required
Your generated audio will appear here
Mirelo SFX 1.6 generates sound effects and seamless ambience loops from English text prompts — whooshes, clicks, sparkles, rain, café noise, room tone. Prizmad ships it as a built-in AI Studio tool priced from 1 token, with up to 4 variations per generation.
Mirelo SFX 1.6 is a text-to-sound model with two distinct modes. Sound Effect mode creates one-time sounds for actions and moments — a cinematic whoosh into a product reveal, a click, an impact, a sparkle chime on a beauty close-up. Looping Ambience mode creates seamless background atmospheres — rain, city noise, café ambience, wind, room tone — built to loop cleanly under a scene without an audible seam.
You describe the sound in English (4 to 2,500 characters), set a duration anywhere from 1 to 60 seconds with a slider, and optionally request up to 4 variations in one run — each variation arrives as a separate audio result, so you can audition takes side by side. Ambience mode adds a Double loop output switch that plays the loop twice in one file, letting you hear whether the seam is actually seamless before you commit it to an edit. Generations run asynchronously with an estimated time of about one to four minutes.
On Prizmad, Mirelo SFX 1.6 completes the audio stack next to Text to Speech and — voiceover, music bed, and now the sound-design layer, all billed from . Generated audio lands in your asset library alongside your video clips, so the whoosh for your product reveal is made in the same workspace as the reveal itself.
A clean, punchy cinematic whoosh timed to the moment your product fills the frame — the single most-used sound in DTC video ads.
Swipes, impacts, and clicks that glue together the rapid scene changes of TikTok- and Reels-style ad edits.
Sparkle chimes for beauty close-ups, clicks for feature callouts, subtle accents that make on-screen text and product details feel tactile.
Seamless rain, street, or café loops under lifestyle footage and AI-generated scenes, so silent generated video gains a believable sense of place.
A low bed of room tone or gentle atmosphere fills the dead air between voiceover lines and keeps the mix from feeling artificially silent.
Sound Effect mode makes one-time sounds for actions, transitions, reveals, clicks, impacts, and sparkles; Looping Ambience mode makes seamless background atmospheres like rain, city noise, or room tone.
Describe the sound in English, from 4 up to 2,500 characters — the workspace examples show the style: "Cinematic whoosh into a bright product reveal, clean polished ad sound, no music, no voice."
A duration slider covers everything from a half-beat click to a full minute of ambience, and the token cost updates automatically as you move it.
Request 1-4 samples in a single generation; each appears as a separate audio result so you can pick the take that sits best in your edit.
In ambience mode, the Double loop output switch renders the loop twice in one file, so you can hear the seam point before dropping it into a video.
Sits alongside Prizmad's Text to Speech and Background Music, giving your ads the sound-design layer — transitions, reveals, atmosphere — that separates polished creative from silence between voiceover lines.
Open the Mirelo SFX 1.6 workspace in AI Studio.
Pick a mode: Sound Effect for a one-time sound (transition, reveal, click, impact) or Looping Ambience for a seamless background atmosphere (rain, city, café, room tone).
Describe the sound in English — 4 to 2,500 characters. Be concrete about character and context, and add "no music, no voice" if you want a clean effect, as the workspace examples do.
Set the duration slider between 1 and 60 seconds; the token cost updates automatically as you adjust it.
Optionally open the advanced controls to request up to 4 variations — note that cost multiplies per variation — and, in ambience mode, enable Double loop output to audition the loop seam.
Click Generate. Results arrive in roughly one to four minutes, each variation as its own audio file in your asset library.
Pick the best take and layer it into your ad in the video wizard, alongside your voiceover and background music.
Cost is 1 token per 5 seconds of audio, rounded up, with a 1-token minimum — a 3-second whoosh is 1 token, the default 10-second setting is 2 tokens, and a full 60-second ambience is 12 tokens. The workspace updates the cost automatically as you move the duration slider.
Requesting multiple variations multiplies the duration cost by the sample count: four 10-second variations cost 8 tokens (2 tokens each). Use variations when you need options for a key moment, a single sample when you just need the sound.
Mirelo SFX 1.6 draws from the same Prizmad token balance as Text to Speech, Background Music, and your video tools — no separate sound-library account, with one-off top-ups available if you run out mid-project.
No — it's a token-based tool on Prizmad. Pricing is 1 token per 5 seconds of audio (rounded up, minimum 1 token), multiplied by the number of variations you request. Short effects are cheap: a 5-second whoosh with one sample costs a single token.
The default settings — 10 seconds, 1 variation — cost 2 tokens. A 3-second effect is 1 token, a 30-second ambience is 6 tokens, and the maximum 60 seconds is 12 tokens. Requesting up to 4 variations multiplies that per-duration cost by the sample count, and the workspace shows the exact number before you generate.
Sound Effect mode creates a one-time sound for an action, transition, product reveal, click, impact, or sparkle. Looping Ambience mode creates a seamless background atmosphere — rain, city noise, café ambience, wind, room tone — designed to repeat cleanly. Ambience mode also unlocks the Double loop output switch, which plays the loop twice in one file so you can verify the seam sounds seamless.
Anywhere from 1 to 60 seconds, set with a slider in 1-second steps. Token cost scales with the duration you choose and updates automatically in the workspace.
Yes — Mirelo SFX 1.6 takes English text prompts, from 4 to 2,500 characters. Concrete, contextual descriptions work best, and adding "no music, no voice" keeps the output a clean effect rather than a musical cue.
Yes. Sound effects and ambience generated on Prizmad are yours to use in paid ads, video content, and product pages.
The workspace estimates one to four minutes. Generations run asynchronously, and if you requested multiple variations they arrive together, each as a separate audio result.
Background Music generates musical tracks — melody, instrumentation, mood — to run under your whole ad for 1 token per track. Mirelo SFX 1.6 generates non-musical sound design: individual effects and environmental loops priced by duration. Most finished ads use both: a music bed from Background Music plus reveal and transition effects from Mirelo.
Published 2026-07-16 · Last updated 2026-07-16