Suno v4

Suno v4 is the fourth-generation music-generation model from Suno, released in November 2024, capable of generating complete songs including vocals, instrumentation, and lyrics from a text prompt in seconds. It is available through the Suno web application at suno.com, the iOS and Android mobile apps, and a limited-beta developer API. As of April 2026, Suno v4 represents the principal step-change release that made AI-generated full songs commercially viable at scale for amateur songwriters and content creators, and Suno remains the leading music-generation platform globally by subscriber count and revenue.

At a glance

• Lab: Suno
• Released: November 2024
• Modality: Audio (music generation: vocals, instrumentation, and lyrics from text prompt)
• Open weights: No (closed)
• Max song duration: Approximately 4 minutes per generation (extendable with the Extend feature)
• Pricing: Free tier (50 credits per day, non-commercial use); Pro ($8/month, 2,500 credits/month, commercial license); Premier ($24/month, 10,000 credits/month, priority generation, commercial license)
• Distribution channels: suno.com web application, iOS app (App Store), Android app (Google Play), Discord integration (early-stage), developer API (limited beta)

Origins

Suno was founded in 2022 in Cambridge, Massachusetts by Mikey Shulman, Georg Kucsko, Martin Camacho, and Keenan Freyberg, all of whom were former engineers at Kensho Technologies, the financial-data AI firm acquired by S&P Global. The founding team's background in audio signal processing and machine learning from the Kensho period informed the technical direction: building foundation models for audio generation rather than adapting general-purpose language or vision models.

Before the consumer Suno product, the founding team released two open-source audio projects. Bark, released in 2023, was a text-to-audio model capable of generating speech, music, and other sounds. Chirp was an earlier internal audio representation model. These projects established the team's public profile in audio AI and informed the architecture underlying the Suno consumer product.

The Suno product launched publicly in late 2023 and achieved rapid consumer adoption through 2024. The version lineage progressed as follows: Suno v1 (2023 public launch), Suno v2 (early 2024, improved audio quality and coherence), Suno v3 (mid-2024, longer generation windows and expanded genre coverage), and Suno v4 (November 2024). Each successive version improved vocal realism, instrumentation clarity, and overall song coherence, but v4 was the release that industry observers and the creator community most consistently identified as the step-change.

Legal context is central to Suno's profile. In June 2024, the Recording Industry Association of America (RIAA) filed a copyright-infringement lawsuit against Suno on behalf of Universal Music Group, Sony Music Entertainment, and Warner Music Group. The lawsuit alleged that Suno had stream-ripped recordings from YouTube and other platforms to harvest training data without authorization. Suno publicly responded by characterizing its training practices as fair use, arguing that AI training on copyrighted material falls under transformative-use doctrine. Warner Music Group subsequently settled with Suno and established a business partnership; the Universal Music Group and Sony Music Entertainment litigation continues as of April 2026.

Capabilities

Suno v4 generates complete songs from a text prompt. A typical prompt describes genre, mood, instrumentation, tempo, and thematic content; the model produces a finished track with vocals, backing instrumentation, and lyrics that match the prompt. Generation takes roughly ten to thirty seconds for a full song.

Vocal realism is the most-cited v4 improvement over v3. Prior Suno versions produced recognizable but frequently synthetic-sounding vocals, with artifacts in consonants, breath transitions, and phrase endings. Suno v4 substantially reduced these artifacts, producing vocals that community evaluations described as approaching the quality of lightly produced demo recordings. The improvement applied across a range of vocal styles from pop to rap to musical-theater belting.

Instrumentation quality and breadth are strong points of the model. Suno v4 handles a wide genre range, including but not limited to pop, rock, hip-hop, country, electronic, jazz, classical-adjacent, metal, R&B, reggae, and ambient. Within each genre, the model generates idiomatic instrumentation: a country track includes pedal steel and fiddle phrasing; a trap track includes 808 bass and hi-hat patterns; a jazz track includes walking bass and chord voicings appropriate to the style. Arrangement coherence, meaning the extent to which instruments complement rather than clash with each other, improved substantially in v4 relative to v3.

Song structure generation is a distinctive capability. Suno v4 produces songs with recognizable verse-chorus-bridge structure rather than undifferentiated audio sequences. The model inserts dynamic changes at structurally appropriate moments, including pre-choruses, instrumental breaks, and outro passages. Lyric quality follows thematic prompts closely, with rhyme schemes and syllabic patterns generally matching the genre conventions of the requested style.

Beyond primary generation, Suno provides several editing and extension features. The Extend feature adds additional sections to a generated track, allowing a short clip to grow into a four-minute or longer song while maintaining style and key. Custom Lyrics mode allows users to supply their own lyrics and have the model generate matching vocals and instrumentation. The Cover feature applies a new vocal or instrumental style to existing audio. Structural tags such as "[verse]", "[chorus]", and "[bridge]" in the lyrics field give users additional control over song layout.

Benchmarks and standing

Music generation has no standardized benchmark infrastructure comparable to the benchmarks used for text or image models. There is no widely accepted equivalent to LibriSpeech, LMArena, or HumanEval for evaluating generated songs. Evaluation is substantially qualitative, relying on creator-community side-by-side comparisons, listener preference studies, and production-viability assessments.

Within those qualitative evaluations, Suno v4 has consistently placed ahead of or comparably to Udio v1.5, the closest direct peer, in listener preference tests conducted by music technology publications and creator communities through late 2024 and into 2025. The evaluations typically ask listeners to prefer one track over another on axes including vocal realism, instrumental quality, production polish, and lyric coherence. Results have varied by genre, with Suno v4 performing particularly well on pop and hip-hop styles and Udio showing stronger results in some experimental electronic and ambient genres.

Suno's commercial scale provides an indirect indicator of standing. As of February 2026, Suno reported approximately 2 million paid subscribers and approximately $300 million in annualized recurring revenue. Industry coverage has consistently characterized Suno as the leading AI music-generation platform globally by both subscriber count and revenue, with Udio as the principal direct peer at smaller scale.

Audio quality metrics such as Frechet Audio Distance (FAD) exist in the research literature but are not standardized across industry evaluations in a way that yields reliable cross-model comparisons. Numbers cited here reflect publicly reported community evaluations as of April 2026.

Access and pricing

Suno v4 is the default model in the Suno product as of late 2024. Access is available through three subscription tiers.

The Free tier provides 50 credits per day, with one song generation consuming 10 credits, yielding 5 full generations daily. Free-tier generations are subject to non-commercial use terms and may be deprioritized during high-load periods. No payment or account tier is required beyond account registration.

The Pro tier is priced at approximately $8 per month (billed annually) or $10 per month (billed monthly) and provides 2,500 credits per month, priority generation, and a commercial-use license. Pro subscribers can also download their generated tracks and disable the Suno watermark.

The Premier tier is priced at approximately $24 per month (billed annually) or $30 per month (billed monthly) and provides 10,000 credits per month, the highest generation priority, and the commercial-use license. Premier subscribers additionally receive access to new features and model variants on an early-access basis.

The suno.com web application is the primary access point with the full feature set. iOS and Android apps offer a subset of the web feature set optimized for mobile. Discord integration is in an early-stage rollout. A developer API in limited beta provides programmatic access to generation endpoints.

Comparison

The peer set for Suno v4 in the music-generation category:

• Udio. The closest direct competitor. Udio was founded by former Google DeepMind researchers and released its music-generation product in April 2024. Udio and Suno were named in the same RIAA lawsuit in June 2024. In qualitative evaluations, Udio v1.5 and Suno v4 are considered the two leading AI music-generation products as of early 2026. Udio has shown particular strength in experimental and ambient genres, while Suno leads on pop and hip-hop styles in community evaluations. Udio's subscriber base and revenue are not publicly disclosed but appear substantially smaller than Suno's reported figures.
• ElevenLabs Music. ElevenLabs entered the music-generation space in 2025 with a music-generation offering distinct from its established voice synthesis and transcription products. ElevenLabs Music is positioned as a newer entrant in full-song generation. ElevenLabs' primary strength in the audio category remains its voice synthesis platform (Eleven v3 and its predecessors), which addresses voice cloning and text-to-speech rather than full music generation. The two products are complementary for creators combining AI-generated voice and AI-generated music in production workflows.
• Stability AI Stable Audio. Stability AI's Stable Audio product offers music and sound generation with open-weights variants available. Stable Audio 2.0, released in 2024, supports generation of tracks up to three minutes in length and emphasizes sound-design and instrumental generation rather than full vocal-and-lyric song production. Stability AI's open-weights positioning differentiates Stable Audio for researchers and developers requiring self-hosted deployment, though commercially the product has not achieved the scale of Suno or Udio.
• Google Lyria. Google's Lyria model, developed at Google DeepMind, surfaces through Google's MusicFX and YouTube Dream Track products. It is not available as a standalone API for general developer access as of April 2026 and is not positioned as a direct creator-platform competitor in the same market segment as Suno.

Outlook

Open questions shaping Suno v4's trajectory through 2026 and beyond:

• Suno v5 release and capability. Suno has publicly referenced a v5 model. The At a Glance section of the Suno lab profile lists both V4 and V5 as current model variants. Whether v5 will be a step-change comparable to the v3-to-v4 transition, or an incremental quality improvement, will determine how the competitive gap with Udio and newer entrants evolves.
• RIAA lawsuit outcome. The Universal Music Group and Sony Music Entertainment copyright-infringement litigation continues as of April 2026. An adverse outcome could constrain Suno's ability to train on recorded music, forcing a shift to licensed or synthetic training data. The Warner Music Group settlement-and-partnership model, if extended to Universal and Sony, would represent a different path: commercially structured access to catalog for training rather than litigation resolution.
• Music-industry licensing model emergence. The Warner partnership may signal a broader structural shift toward licensing arrangements between AI music-generation companies and major labels. If licensing norms emerge, they could change the competitive landscape for smaller music-generation entrants who cannot afford label licensing deals at Suno's scale.
• Competitive pressure from new entrants. ElevenLabs' entrance into full-song generation, alongside Google Lyria's distribution reach through YouTube and potential expansion, could intensify competition beyond the current Suno-Udio duopoly. The pace at which new entrants match v4-quality vocal generation will be a key competitive variable.
• Commercial licensing clarity for creators. Pro and Premier tiers grant commercial-use rights, but how streaming platforms and collecting societies handle AI-generated music at scale remains contested and will shape the practical monetization path for professional creators using the platform.

Sources

• Suno official site. Product reference and subscription tier pricing.
• Suno v4 announcement blog post. Release notes and capability overview for the November 2024 v4 launch.
• Time: Major Record Labels Sue AI Music Generators. RIAA lawsuit coverage, June 2024.
• TechCrunch: AI music startup Suno claims training model on copyrighted music is 'fair use'. Suno's public fair-use response to the RIAA complaint.
• TechCrunch: AI music generator Suno hits 2M paid subscribers and $300M in annual recurring revenue. February 2026 commercial scale reporting.
• WebProNews: Suno's CEO Calls AI Prompts 'Active' Music Creation Amid $2.45B Valuation Boom. Valuation and CEO framing of the product.
• Resemble AI: In-depth details About the Lawsuits Against Suno and Udio by RIAA. Litigation detail covering both Suno and Udio complaints.