Introducing Mirelo AI Sfx V1.5 Video-to-Video on WaveSpeedAI
Try Mirelo AI Sfx V1.5 Video-to-Video for FREEMirelo SFX V1.5 Video-to-Video Is Now Available on WaveSpeedAI
The world of AI-generated video has a silent problem—literally. While models like Sora, Veo, and Kling have revolutionized visual content creation, they’ve left creators with stunning footage that lacks the sonic dimension that brings media to life. Today, we’re excited to announce that Mirelo SFX V1.5 is now available on WaveSpeedAI, giving you the power to transform silent videos into fully synchronized audiovisual experiences.
What Is Mirelo SFX V1.5?
Mirelo SFX V1.5 is a cutting-edge video-to-audio model developed by Berlin-based Mirelo AI, a company founded by former AWS Labs researchers CJ Simon-Gabriel and Florian Wenzel. Both founders bring unique credentials to the table: CJ holds a PhD in machine learning from the Max Planck Institute with a postdoc at ETH Zurich, while Florian earned his PhD in deep learning from Humboldt University and previously worked at Google Brain.
The model uses advanced multimodal AI to analyze video content and generate perfectly synchronized sound effects. It doesn’t just detect motion—it understands context. Whether your video features footsteps on gravel, rain hitting windows, or dramatic explosions, Mirelo SFX V1.5 creates realistic, cinematic-quality audio that matches the visual rhythm of your content.
Key Features
AI-Driven Sound Synthesis
The model generates sound effects that precisely match object motion, timing, and energy directly from video frames. Unlike simple audio overlays, Mirelo’s approach ensures every sound corresponds to what’s actually happening on screen.
Cinematic Awareness
Mirelo SFX V1.5 detects on-screen actions including impacts, motion intensity, and scene transitions, producing effects that feel professionally crafted. The model understands the difference between a gentle tap and a forceful strike, adjusting audio characteristics accordingly.
Superior Quality in Blind Tests
In independent evaluations, Mirelo SFX V1.5 achieved a 68.3% win rate (excluding ties) and 73.2% (including ties) when compared against popular alternatives like Kling Text-to-Audio and Tencent-Hunyuan VideoFoley. Users preferred Mirelo’s output 67-77% of the time in listening tests.
Production-Ready Output
The model delivers clean, contextual sound effects without the audio artifacts, distortion, or unwanted music/speech leakage that plague many competitors. What you get is ready for professional use.
Lightweight and Fast
Mirelo’s architecture requires 50 times less compute than typical large language models while still delivering superior quality. Generation happens at up to 1.7× faster than real-time, meaning a 10-second video can have its sound effects generated in roughly 6 seconds.
Multiple Variations
Generate multiple sound versions for the same video, giving you creative control during post-production. Audition different takes before selecting the perfect audio for your final cut.
Real-World Use Cases
Content Creators and Social Media
Transform your AI-generated videos from silent clips into engaging content. Whether you’re creating TikToks, YouTube shorts, or Instagram Reels, synchronized audio dramatically increases viewer engagement and watch time.
Film and Animation Production
Speed up post-production workflows by automatically generating foley sounds. While professional Foley artists remain invaluable for hero moments, Mirelo SFX V1.5 can handle background audio and secondary sound effects, freeing up resources for creative work that matters most.
Game Development
Prototype audio for game cinematics and cutscenes rapidly. Generate placeholder sounds that communicate the intended experience to stakeholders before investing in custom audio production.
Marketing and Advertising
Create polished video ads without expensive sound design sessions. E-commerce brands can produce product videos with appropriate ambient audio, while agencies can iterate faster on creative concepts.
AI Video Enhancement
If you’re using AI video generators like Sora, Veo, Kling, or Wan, Mirelo SFX V1.5 serves as the perfect companion. Generate your visuals, then add synchronized audio in seconds—completing the audiovisual experience in a single workflow.
Getting Started on WaveSpeedAI
Using Mirelo SFX V1.5 on WaveSpeedAI is straightforward:
- Upload your video via drag-and-drop or paste a URL (supports MP4, MOV formats)
- Add an optional prompt describing the sound context (e.g., “soft footsteps on wood,” “metal clangs,” “rainy street ambience”)
- Set the number of samples to generate multiple variations for creative flexibility
- Click Run and receive synchronized audio in seconds
The model processes videos up to 10 seconds in length, with typical generation times of 6-12 seconds per run. For best results, use short, focused clips with clear, high-contrast motion.
Pricing
Mirelo SFX V1.5 offers predictable, affordable pricing:
- 0-5 seconds: Minimum charge applies ($0.035 × number of samples)
- 5-10 seconds: Billed by actual duration ($0.007 × samples × duration)
- Maximum per run: $0.07 × number of samples
Pro Tips for Best Results
- Use clips under 10 seconds with focused action for strongest visual-sound alignment
- Include contextual prompts like “rainy street, distant thunder” for more nuanced results
- Generate 3-5 samples to audition variations before selecting your final audio
- Adjust the seed value for subtle timing and tonal variations while maintaining synchronization
Why WaveSpeedAI?
When you run Mirelo SFX V1.5 on WaveSpeedAI, you benefit from:
- No cold starts: Your requests process immediately without waiting for model initialization
- Fast inference: Optimized infrastructure delivers results quickly
- Affordable pricing: Pay only for what you use with transparent per-second billing
- Simple API integration: Integrate video-to-audio capabilities into your applications with our REST API
The Future of Audiovisual AI
The release of Mirelo SFX V1.5 represents a significant milestone in closing the audio gap that has limited AI-generated video content. Backed by a recent $41 million seed round from Index Ventures and Andreessen Horowitz, Mirelo continues to push the boundaries of what’s possible in AI sound generation.
As AI video models become increasingly sophisticated, the demand for synchronized audio will only grow. Mirelo SFX V1.5 positions creators to stay ahead of this curve, transforming silent AI videos into complete multimedia experiences.
Start Creating Today
Ready to bring your silent videos to life? Mirelo SFX V1.5 is available now on WaveSpeedAI. Experience the difference that perfectly synchronized, AI-generated sound effects can make in your content.
