ACE Music v1 Tops Artificial Analysis Vocal Music Board at 1160 Elo, Within Margin of Suno v6
ACE Music v1 ranks first on Artificial Analysis' AA-Music-Vocal v1.1 leaderboard with an Elo of 1160 (±21), ahead of Suno v6 at 1144 (±18). The 16-point gap falls inside overlapping 95% intervals, so the board does not separate the two with confidence.
ACE Music v1, a music generation model from the team behind the ACE Studio AI music workstation, ranks first on Artificial Analysis' AA-Music-Vocal v1.1 leaderboard with an Elo score of 1160 (±21), ahead of Suno v6. The lead is small and within the margin of error.
Leaderboard standings
Pandaily checked the live leaderboard at about 10:00 Beijing time on November 16. Figures below reflect that reading and may have changed since.
| Model | Elo |
|---|---|
| ACE Music v1 | 1160 (±21) |
| Suno v6 | 1144 (±18) |
| Suno v6-mini | 1132 (±18) |
| Mureka V9.5 | 1101 |
| Mureka V9 | 1063 |
| Lyria 3.5 | 1050 |
ACE Music v1's score comes from 2,119 samples. Each of the other leading models has more than 3,000.
The 16-point gap between ACE Music v1 and Suno v6 sits inside overlapping 95% intervals. Artificial Analysis gives ACE Music v1 a rank range of 1 to 2 and Suno v6 a range of 1 to 3, so the board does not separate the two with confidence.
ACE Music v1 does not appear on the separate instrumental leaderboard, where Suno v6 leads at 1148.
How the ranking works
The board is built from blind comparisons. Evaluators hear two tracks generated from the same prompt without knowing which models produced them, then pick the one they prefer. ACE notes that the ranking reflects overall vocal music preferences, not performance on separate genres or features.
Model features
ACE Music v1 takes text descriptions and lyrics as input. Users can add performance cues in parentheses, such as "hum softly." Songs can be exported to MIDI and split into stems.
A separate ACE Music v1.1 Preview plans music in ABC notation before generating audio, so users can edit a score and regenerate. The company plans to open-source a lighter model, ACE Music v1 Flash, in late 2026.
Pricing and access
According to Chinese tech outlet QbitAI, free users of the ACE Music app get 20 credits a day, enough for up to 10 songs. The Creator plan costs $9.90 a month. ACE says songs made on paid plans carry commercial use rights. API pricing is not yet disclosed.
Background
The team began developing its music foundation model in March 2025 and also maintains the open-source ACE-Step music models. Its acemusic.ai site lists Timedomain, Inc. as the operator.
What this means
The headline rank is real but not decisive. With overlapping confidence intervals and rank ranges of 1 to 2 for ACE Music v1 and 1 to 3 for Suno v6, the data supports calling the two close, not calling ACE Music v1 the clear leader. The lower sample count for ACE Music v1 (2,119 against more than 3,000 for the other leading models) is a further reason to read the score cautiously.
The result also covers only vocal music, judged on overall listener preference. ACE Music v1 has no entry on the instrumental board, and the ranking says nothing about individual genres or features such as MIDI export and stem separation.
Related Articles
Open-weights model ranks in top 3 of AI video Elo ranking as native audio becomes standard
MarkTechPost's AI video generation ranking, dated Nov 13, 2026, uses blind human-preference Elo from the Artificial Analysis Video Arena. An open-weights model now sits in the top 3, and native audio is described as table stakes.
Gemini 4 Hydrogen scores 54 on Artificial Analysis Intelligence Index, just below Claude Sonnet 5.5 and GPT-6.1 Sol
Artificial Analysis reports that Google's Gemini 4 Hydrogen scores 54 on its Intelligence Index, just below Claude Sonnet 5.5 and GPT-6.1 Sol. It costs a third of Sonnet's cost but about 2.5x that of GPT-6.1 Sol, and adds an xhigh reasoning level.
Claude Fable 5.2 paints images through Python code, using about 700,000 simulated strokes in one piece
Anthropic's Claude Fable 5.2 produces images by writing Python programs that drive a simulated paint engine, with no external image model. Anthropic reports 64.9 percent on its internal RoboDojo Sim benchmark against 29.8 percent for OpenAI's GPT-6 Astra, though a real-world robot arm test showed limits in speed and fine motor control.
Gemini 4 Carbon tops MathArena at 85.4%, ahead of Claude Fable 5.2 and GPT-6.1 Sol
Google's Gemini 4 Carbon (high) leads the MathArena math benchmark with an expected performance of 85.4%, ahead of Anthropic's Claude Fable 5.2 (max) at 82.8% and OpenAI's GPT-6.1 Sol (max) at 80.6%. With the reasoning settings used in testing, Carbon costs roughly 15 times as much as GPT-6.1 Sol.
Comments
Loading...