Biniôke
Karaoke for the AI age. Word-for-word lyric sync via WhisperX, AI-curated backgrounds bound deterministically per track, optional typography overlays, dual-vocal split. Built and used internally — public release blocked on the right billing infrastructure.
01 What it is
Drop in a song (file, link, or Suno track), and Biniôke produces a karaoke-ready video: lyrics that highlight word-by-word in sync with the audio, an AI-generated visual backdrop bound deterministically to that specific track (so the same track always gets the same visuals — reproducible, not random), and your choice of typography overlay. Optional dual-vocal split separates lead from harmonies so you can sing one line and the other plays back.
Word-level sync via WhisperX
Not line-by-line. Each word gets its own timestamp from forced-alignment, so the highlight tracks the actual vocal — even for fast rap or held vowels.
Deterministic AI visuals
The visual seed is derived from the track itself. Same track in = same visuals out. Reproducible across renders, no slot-machine vibes.
Typography modes
Five text-effect modes × four letter-styles × three sizes × 23 visualizer modes. Combinable per render. Pick a vibe, lock it.
Dual-vocal split
Separates lead from harmony stems via Demucs. Sing one, hear the other. Useful for rehearsal, duet performance, vocal practice.
02 Why this is not for sale yet
Honest version: Biniôke is built. It's been used internally on hundreds of tracks. The reason it's not on the buy-page is structural, not technical:
What's blocking public release
- Recurring billing system not yet operational. Biniôke's image-generation per render has real per-track API cost. A €5-once price would lose money on heavy users; a fair tier needs monthly billing, which our SEPA-based pipeline does not yet support natively.
- Belgian KBC business account activation pending. Until that's live, we cannot legally process recurring B2C subscriptions in the EU under our own name.
- Image-generation key requirement. Free Gemini tiers will not cover production-scale image generation; the launched version requires either a built-in paid pool (subscription model above) or BYOK with OpenAI/equivalent.
None of these are model-quality problems. They are payment-infrastructure problems. Once the KBC + recurring stack is live, Biniôke launches publicly within days.
03 The plan when it ships
Tiered pricing, locked when the billing infra is in place. Current canonical pricing (subject to small adjustments at launch):
• Demo — free, browser-WASM Whisper, watermarked output
• Basic — €5 / mo · standard renders, web-app access
• Pro — €25 / mo · faster renders, all visual modes, dual-vocal split, no watermark
• Suno-tier add-on — €8 / mo · auto-pulls from Suno playlists via Sunôgrâtor integration
• Ultimate — €49 / mo · everything + priority queue
• Venue licensing — €99 / €299 / €599 monthly tiers for bars / event spaces / chains
• Performer revenue-share — 10–15% on covers monetized via Biniôke output, when the listener consents
If the per-render image-gen cost ever drops dramatically, the prices come down with it — that's locked in writing.
04 Get notified at launch
Drop your email at the top of this page. One notification when Biniôke opens to the public — nothing else, no newsletter. If you want to test it earlier as a closed-beta participant: email beta@biniruprojects.ai with a short note on what you'd use it for.