The 10 Best AI Tools for Sound Effects in 2027

ElevenLabs turns a text prompt into four sound effects at a time, each up to 30 seconds, with a seamless looping option. Downloads come as MP3 (44.1kHz) or WAV (48kHz). Free is $0 with 10k credits and 50 generations a month for personal use only; Starter is $6 a month with 30k credits and 150 generations, Creator is $22 and Pro is $99. Paid accounts can use the effects royalty-free in commercial projects.

Adobe Firefly’s sound effect generator lets you type a prompt or act the sound out into your mic so the timing and intensity match your video. Each generation produces 4 variations of up to 30 seconds, and prompts are English only; the feature is in beta in the Firefly video editor timeline. Firefly Standard is US$9.99 a month with 2,000 credits and up to 200 sound effects; Pro is US$19.99 with up to 400. Adobe says every effect is royalty-free and commercial-ready.

Krotos Studio lets you perform Foley, ambiences and whooshes in real time with a mouse, keyboard or MIDI controller, and its Video to Sound analyses footage to generate synchronised audio in Premiere Pro, DaVinci Resolve or the browser. Studio costs $19.99 a month billed monthly or $9.99 billed annually and includes >33,000 WAV files; Pro is $49.99 billed monthly and Max is $83.33 a month on yearly billing. A 5-day free trial runs on macOS and Windows.

Stable Audio, from Stability AI, makes sound effects up to 30 seconds and music up to 6 minutes, and every plan, free included, comes with a commercial license. Free is $0 with 50 credits a month; Solo is $12 a month with 660 credits, then Session $30, Producer $90 and Studio $199. All plans include the web app and the DAW plugin.

SFX Engine is a dedicated sound effect generator for ambience, foley, impacts, whooshes, UI sounds and creatures up to 47 seconds long, with up to four takes side by side. You can try 5 sounds free; Starter is $10 a month for 500 credits, at 3 credits per sound, with Creator $25 and Pro $90. Every sound is royalty-free for games, videos, ads, podcasts and apps.

Mirelo generates sound for an uploaded video: footsteps, impacts, foley and ambience that land where they should, for up to 10 minutes in one video with dialogue included. Free is 0 € a month with 5,000 credits and a watermark on exports; Creator is 20 € a month with 24,000 credits and Studio is 99 €. It has plugins for Adobe Premiere Pro, DaVinci Resolve and Roblox, plus an API.

Fish Audio’s sound effects generator creates effects from text descriptions and lets you fine-tune duration, guidance and steps. It sits inside a voice platform whose Free Tier is $0 with 8,000 credits monthly; Plus is shown at $5.5 a month in the first year with 50% off, $66 billed annually, and lists commercial use allowed.

CyberLink’s MyEdit has a browser-based text-to-sound tool: describe a sound in up to 800 characters, pick a duration and download it as MP3, WAV, FLAC or M4A. There is a free version with limited use and 3 daily free credits; Audio Pro is $5 a month billed $60 yearly with 360 credits a month.

HunyuanVideo-Foley from Tencent Hunyuan is an end-to-end text-video-to-audio model that generates Foley synced to video at 48kHz. Its inference code and model weights are publicly available, and the XL model needs 16GB VRAM, or 8GB with offload. It ships under the Tencent Hunyuan Community License, which does not apply in the European Union, United Kingdom and South Korea.

Meta AudioGen is the text-to-sound model in the open-source AudioCraft code base, trained on public sound effects. The released facebook/audiogen-medium is a 1.5B model and needs a GPU with at least 16 GB of memory. The code is MIT-licensed, but the model weights are CC-BY-NC 4.0.