ZeroHour
Story · 3 sources · 3 articlesfirst updated ()1

Suno launches v6 AI music model family trained on licensed music from Warner, BMG, and Believe as Sony and Universal suits continue

infoModel releaseimportance 55
What's new: Since the previous summary (2026-09-09T14:17:55Z), The Verge's later report (21:42 UTC) adds: v6 was trained from the ground up on a new dataset that also includes user data; it is unclear whether all dubiously obtained content was excluded; the reviewer saw dramatically improved genre fidelity; v6 still cannot produce intentional imperfections like off-key or dissonant vocals; and rollout starts…
Merged summary · glm-5.3-flash · rewritten as coverage arrives

Suno released its v6, experimental v6-wild, and free v6-mini music models — its first built with licensed data from Warner Music Group, BMG, and Believe — adding plain-language editing of song sections and multimodal prompts, while litigation from Sony,…

Suno launched a new AI music model family — v6, experimental v6-wild, and v6-mini — its first built with record industry involvement, trained from the ground up (The Verge) on a new dataset that includes licensed music from Warner Music Group, BMG, and Believe plus user data; The Verge notes it is unclear whether all dubiously obtained content was excluded. The Decoder reports the models were built with Warner, BMG, and Believe, with flagship v6 serving Pro and Premier subscribers, v6-wild aimed at experimentation ('more unpredictable' per The Verge), and resource-light v6-mini free to all users; TechCrunch describes v6-wild as experimental and v6-mini as faster. Capabilities: plain-language/text-based editing of individual song sections — lyrics, vocals, and mashups — without regenerating the full track (The Decoder, The Verge); generation from text, images, and video references (TechCrunch), with The Decoder and The Verge also citing audio prompts; instrument/stem separation (TechCrunch); understanding of vocals, instrumentation, structure, mood, and multimodal inputs (The Decoder); and a planned opt-in artist remix program (TechCrunch). The Verge's reviewer found dramatically improved genre fidelity but that v6 still cannot produce intentional imperfections like off-key or dissonant vocals. Rollout of v6 starts now, with older models to be retired: TechCrunch says Suno plans to retire older models, The Decoder says all older models are being retired, and The Verge says they will eventually be retired. Legally, the launch follows Warner's November 2025 licensing settlement (The Decoder; TechCrunch says 'last year') and a BMG settlement last month (August 2026, per TechCrunch), while litigation continues from Sony, Universal Music Group, and artists including Jason Isbell (TechCrunch); The Decoder adds that a Munich court found Suno's v3.5 and v4 infringed six works and that the size of the training dataset is kept under seal in the US District Court in Massachusetts. Suno reports more than 100 million users, over two million paying subscribers, and $300 million ARR (The Decoder); TechCrunch says it has raised over $819 million to date, and The Decoder separately reports a $400 million raise at a $5.4 billion valuation in June 2026.

  • Suno launched the v6 model family: flagship v6 (for Pro and Premier subscribers, per The Decoder), experimental v6-wild ('more unpredictable' per The Verge), and free v6-mini (The Decoder, The Verge).
  • Trained from the ground up (The Verge) on a new dataset including licensed music from Warner Music Group, BMG, and Believe, plus user data (The Verge); it is unclear whether all dubiously obtained content was excluded (The Verge).
  • Plain-language/text-based editing of individual song sections — lyrics, vocals, and mashups — without regenerating the full track (The Decoder, The Verge).
  • Generation from text, images, and video references (TechCrunch); The Decoder and The Verge also cite audio prompts. The Decoder says all three models understand vocals, instrumentation, structure, mood, and multimodal inputs.
  • Instrument/stem separation is cited by TechCrunch, which also lists a planned opt-in artist remix program.
  • The Verge's review found dramatically improved genre fidelity, but v6 still cannot produce intentional imperfections like off-key or dissonant vocals.
  • Rollout starts now; retirement of older models differs slightly by source: TechCrunch — plans to retire; The Decoder — retiring all older models; The Verge — eventually retired.
  • Settlements: Warner in November 2025 (The Decoder; TechCrunch says 'last year') and BMG last month, i.e., August 2026 (TechCrunch); a Believe arrangement is cited by all three reports without a date.

Coverage timeline

  1. · 6d ago
    TechCrunch · AI· 52
    Suno replaces its AI models with a new one trained on licensed music as copyright suits pile up

    Suno launched its v6 model family trained on licensed music from Warner, BMG and Believe as copyright lawsuits from Sony and Universal continue.

  2. · 6d ago
    The Decoder· 50
    Suno launches v6 music models built with Warner, BMG, and Believe

    Suno launches v6, v6-wild, and free v6-mini music models co-developed with Warner Music Group, BMG, and Believe, retiring all older models.

  3. · 6d ago
    The Verge · AI· 55
    Suno releases its first AI music model made with record industry help

    Suno released its v6 music model family (v6, v6-wild, v6-mini), the first trained with licensed data from Warner Music Group, BMG, and Believe.