Skywork AI's Mureka V9 Closes Within 1 Elo Point of Suno's Top Spot

Artificial Analysis just published its blind-vote leaderboard results for Mureka V9, the latest flagship music model from Skywork AI. On both the Instrumental and Vocals leaderboards, V9 debuts at #2, directly behind Suno V5.5. In the Instrumental category, the gap is a single Elo point.

How Close Is Close?

The Artificial Analysis Music Arena ranks models using an Elo rating system, the same method used to rank chess players. Real users listen to pairs of AI-generated tracks from the same prompt and vote for whichever they prefer, without knowing which model produced either track. Because there is no self-reported metric to game, it is one of the more reliable benchmarks available.

Current standings:

Suno V5.5 is a formidable target. It produces longer single-pass tracks without stitching artifacts, cleaner stems in its Studio editor, and automatic mixing that reduces the need for post-generation mastering. At one Elo point of separation on instrumentals, the difference between the two models may come down to personal taste rather than any measurable quality gap.

The Architecture Behind the Sound

Mureka V9's most technically interesting feature is MusiCoT, short for Music Chain-of-Thought. The idea borrows from reasoning models in the LLM world: rather than generating audio in a single pass, the model first plans the song's structure, then produces audio tokens. It maps out intro, verse, chorus, bridge, and outro before any audio is generated.

That planning step is why V9 tends to hold coherence across longer tracks. Fewer songs collapse after the first chorus, section contrast is clearer, and transitions feel more deliberate. The key capability upgrades in V9 include:

• Section-by-section control over emotion, instrumentation, and BPM, so you can specify "build tension in the bridge, drop to acoustic guitar in the outro"
• Improved text-to-music fidelity, with vocals that track lyrical intent more closely
• Full songs typically running two to four minutes, in vocal or instrumental modes
• Lyrics in English, Chinese, Japanese, Korean, and Spanish
• Track splitting into up to 12 stems with audio or MIDI export
• Faster generation and more output diversity with less repetition

The DAW Integration That Sets It Apart

Mureka's sharpest competitive edge over every other AI music tool is what happens after generation. It exports music as MIDI files, meaning melodies, basslines, and chord progressions can be pulled directly into Ableton, Logic, FL Studio, or any other DAW. For producers who want AI as a compositional starting point rather than a finished product, that changes the workflow entirely.

Mureka also ships Mureka Co, a desktop app that operates inside your DAW via natural language. Describe a part, and it generates audio, splits stems, extends sections, and imports everything beat-aligned into the session. As of mid-2026, it supports Ableton Live only, with additional DAWs promised. No other major AI music generator offers an agentic in-DAW workflow at this level.

Suno produces excellent finished tracks, but it stops at the export. Mureka's stem separation, MIDI export, and support for over ten languages give it a clear lead for producers who need to take AI output further, particularly for non-Western musical traditions and tonal languages where Suno's results are less consistent.

A Crowded Market at the Top

The competitive landscape has compressed fast. By early 2026, Suno V5 was producing vocals that genuinely fool listeners, Google shipped Lyria 3 with singing capabilities, and ElevenLabs brought its audio expertise to music generation. Mureka V9 is entering the most competitive moment the space has seen.

Pricing across the field is relatively tight. Mureka's Pro tier runs $9/month for most users, with a free tier offering around 50 credits per day and a Premier tier at $27/month for heavier workflows. Suno's free tier also offers 50 daily credits, but those credits run on the older V4.5 model, not V5.5.

The Legal Context

In December 2025, plaintiffs filed Attack the Sound LLC v. Kunlun Tech Co. in the U.S. District Court for the Northern District of Illinois, alleging that Mureka was trained on copyrighted recordings without authorization and that its "royalty-free" marketing is misleading. The complaint includes counts for copyright infringement, DMCA violations, and Illinois biometric and right-of-publicity claims. No liability has been found, but the case is active.

It is the 66th U.S. copyright case filed against an AI company, and only the second targeting a Chinese technology firm in a U.S. court, following Disney v. MiniMax. Every major AI music generator faces the same unresolved questions about training data. Anyone building commercial workflows on Mureka should track the case.

What to Do With This

If Suno is already working for your finished-track pipeline, a one-Elo-point gap is not a reason to switch. But if your workflow involves pulling AI-generated material into a DAW, remixing stems, or exporting MIDI for further production, Mureka V9 is the only major platform that handles all of it in one place.

Mureka V9 is available now for blind comparisons in the Music Arena, on the Vocals Leaderboard, and directly through the Mureka app and API.