mirelo/video-to-sfx-v1.5

Generate synced sounds for any video and return it with its new soundtrack - now enhanced in version 1.5 for improved sound synchronization and realism

29 runs
Playground API Examples README

Mirelo SFX V1.5

Generate synced sounds for any video using Mirelo’s latest model version. This is an updated version of Mirelo’s video-to-sound model with improved performance and quality.

This model takes a silent video as input and unmutes it by generating sound effects (sfx) that are synchronized with the visuals. The result is the original video returned with a realistic, synced soundtrack.

What’s New in V1.5

  • Updated Model: Uses Mirelo’s latest v1.5 model for improved sound generation
  • Enhanced Quality: Better audio quality and more accurate sound synchronization
  • Improved Performance: Optimized generation process for better results

Features

  • Input: Silent video (max 10 seconds). Videos longer than 10 seconds will be truncated.

  • Output: Video with high-quality, synced sound effects.

  • Optional Prompt: Guide the sound generation with text (e.g., “metal clanging” or “birds chirping”).

  • Multiple Variations: Choose between 2–4 output samples per run.

  • Advanced Controls: Fine-tune generation with creativity coefficient, steps, and start offset parameters (same as v1.0).

Generated audio is intended for sound effects and syncing — not speech or music.

Pricing

This model is priced at $0.01 per second of output audio per sample. For example, 2 samples of 10-second audio = 20 units × $0.01 = $0.20.