Music video from your track

Eval results

Scored on a held-out or test splitRun 26 Sep 2026Eval write-up (decosa-api, access required)

• Lyric lines within 0.3 s / 1 s of human timing81.8% / 87.7%test splitn = 66 held-out songs; 4 of 6 songs have at least 80% of lines within 1 s. Baseline 16.8% within 1 s. Dev: 67.0% / 67.9%.
• Beat F-measure (±70 ms), artist's BPM given0.868test splitn = 40Without BPM: 0.591. Drums-only grooves: an upper bound for full mixes.
• Typed check says no to a swapped (mismatched) scene67.3%test splitn = 52The code checks (citations, anchor words) carry the grounding; the typed check lets a third through.
• Anchor words found in the cited lines96.2%test splitn = 52Citations valid: 100%
• Rendered cut offset from the beat grid (median / max)7.0-10.0 ms / 16.3 mssynthetic5 renders of one excerpt; against the detected grid, not a human one. Some cuts between similar shots are not detectable (13-21 of 20-21 found).
• Render cost, one formatabout $0.32 per minute of videosyntheticAt an assumed $1.69/h GPU rental price; the hosted demo does not bill renders.

Nightly smoke check

Loading the nightly status…

Result

pass

Run

26 Sep 2026

Latency, this run

n/a

p50 over passed runs

15 s

Receipts

7

Model calls

n/a

Tokens

n/a

Cost per run

$0.002

Self-host verification

Verified on 26 Sep 2026: Fresh clone of the branch into a clean directory, docker build of the api image and services/mvideo, compose with named volumes on host networking, pointed at the already-running local Qwen (direct route) and ComfyUI; then torn down.

No rights statement: 403; the sample analysed in 19 s (112.35 BPM, 18 lines, 6 grounded scenes, 20 cuts, 7 attested receipts); a 9:16 render finished with a C2PA credential, the disclosure rules met and a record that verified. Found on the way: the containerised analyzer's beat grid sat about 160 ms later than the host's on the same file (different decoder), and a clip of flickering neon fooled the cut detector (fixed: isolated spikes only).

Models and licences

Standard tier. Licence posture: permissive (Apache, MIT or BSD).

• Beat, bar and section detection (CPU): librosa beat tracker on a full-band plus low-band onset envelope, bar phase by a kick-and-snare heuristic (4/4), sections by checkerboard novelty on chroma and MFCC self-similaritydecosa-mvideo-analyze (services/mvideo)AGPL-3.0-or-later (decosa-api)
• Lyric timing (CPU): CTC forced alignment of the artist's own lyrics on the mix, with a repair pass for lines squeezed into too little time; English letterswav2vec2-large-960h-lv60-selfApache-2.0
• Treatment writer (scenes that cite the lyric lines they show) and the typed yes/no grounding check per scene; also labels near-duplicate lyric lines in the clearance pre-checkQwen3.8-27B (NVFP4)Apache-2.0
• Sample and lyric clearance pre-check on the upload (tool 38, run in-process): audio landmarks and melody against a small open catalogue, lyric lines against a lyric setdecosa-api clearance module (decosa_api/verticals/clearance)AGPL-3.0-or-later
• Clips: text-to-video, one 5 s clip per scene per formatWan2.2-VACE-Fun-A14B + Wan2.2-Lightning 4-step LoRAsApache-2.0
• The edit and the marks (CPU): cuts on bar lines on a 30 fps grid, karaoke captions (ASS, libass), the AI label on a top bar, the credit, the artist's audio; cut timing measured back from the pixels; the disclosure pre-flight's rules (tool 49) on the file; C2PA credential per file; the signed recorddecosa-api mvideo module + FFmpeg + c2pa-pythonAGPL-3.0-or-later

All quality evidence

Every sourced number on the tool’s Stack tab, by tier. Some are proxies from another task; their labels say so.