AudioGen vs SFX Studio

AudioGen is a text-to-sound model provided through AudioCraft. It generates audio samples from text descriptions and can continue from an audio prompt, with conditional and unconditional generation modes. Sampling controls include greedy, temperature, top-K, and top-P methods. The listed pretrained model, facebook/audiogen-medium, has 1.5 billion parameters; the example API generates sound and saves it as WAV. The implementation uses an autoregressive Transformer and a 16 kHz EnCodec tokenizer. The repository provides API access and a local Jupyter notebook demo. Running the medium model locally requires a GPU with at least 16 GB of memory; installation requires Python 3.9 and PyTorch 2.1.0, with ffmpeg recommended. The model was trained on English descriptions and performs less well in other languages. It cannot generate realistic vocals, and prompts may need refinement. Code is MIT-licensed, while weights use CC-BY-NC 4.0; training datasets are not provided. The listed self-hosted plan is free.

SFX Studio is an AI sound-effects generator for film, games, trailers, and motion. Users describe what happens on screen and receive a sound-effects take. Listed sound areas include whooshes, impacts, transitions, UI cues, textures, footsteps, cloth, props, abilities, hits, pickups, ambience, and weapon cues. Foley prompts can cover movement, handling, and contact sounds, while game workflows include interface feedback, combat, magic, machines, pickups, ambience, and prototype audio. SFX Studio is the focused front door; the complete generation workflow runs inside TwoShot. New TwoShot accounts receive free credits at signup without a credit card or trial expiration date. The free tier has no watermark or resolution cap. Outputs are available as WAV and MP3, and commercial use is included. TwoShot’s integrations hub covers MCP, REST API, automation, IDE agents, and Coproducer. Plans are Free at free, Pro at $9.50 a month, and Credit pack at $15, paid once. Outputs cannot be directly resold, redistributed, or used to train models.