Agents/YuE
74/100

YuE

Evidence review 3.8/5

Open-source music large model for full-song generation, from lyrics to complete songs

Which tools fit your actual task?

Describe the task and required capabilities, then compare catalog candidates by budget and language. Verify task performance separately.

Select tools for my task →
TL;DR

At a glance

YuE2 is an open-source music foundation model for lyrics-to-full-song generation. It can generate vocals and accompaniment simultaneously, supports multilingual singing, and introduces symbolic planning, zero-shot cover singing, and agentic music editing capabilities. It provides locally deployable inference code and weights, and is a generative AI application that can be run directly, rather than a component only for integration. For teams that want to do AI music creation or secondary development, it is a rare end-to-end solution in the current open-source camp.

Strong open-source control and value, but speed and Chinese quality need real testing.

Catalog information score 74/100 is not measured task performance. The separate evidence review scored 3.8/5.

Verification and change timeline

Only public append-only events appear here; automated and human verification are labelled separately.

  1. Review sources checked

    Automated review published · YuE

    An automated review checked YuE2 against public sources and structured scenarios.

    Source evidence checked
    Check source ↗
  2. Review sources checked

    Automated Evaluation Released · YuE

    The automated evaluation system cross-checked YuE2 and structured scenarios against public sources.

    Source evidence checked
    Check source ↗
  3. Published

    Official version information updated · YuE

    The version model appears verbatim in the official website evidence captured this time, and the system has automatically updated the catalog.

    Source evidence checked
    Check source ↗
  4. Discovered

    Newly discovered and cataloged · YuE

    Scout discovered it from public sources; it entered the catalog after passing the automated publishing quality gate and does not yet equal trusted hands-on testing.

    Source observed
    Check source ↗
Evidence review

Review findings

This is a source review, not a measured task run. The recorded product/version does not identify an executed model. Manual publication alone does not establish hands-on testing; inspect benchmark runs for actual models, tasks and dates.

Based on public information: YuE2 is an open-source music foundation model that turns lyrics into a complete song, generating vocals and accompaniment together and supporting multilingual singing. Unlike most open-source options that output instrumentals only, it adds symbolic planning, zero-shot covers and agentic music editing, giving more control over song structure and revision — hence 4.0 for task completion, with the caveat that release-grade audio quality is unverified. Quality scores 4.0 for controllability rather than claimed fidelity. Chinese support is only 3.5: the site says multilingual singing but offers no Chinese-specific tuning or samples. Value is the standout at 4.5 — open weights, local deployment, no per-call fees. Speed gets 3.0 because no latency or full-song generation time is published and local inference is compute-heavy. Verdict: best for creators who self-host and teams building on top; less suited to casual users wanting plug-and-play.

Strong open-source control and value, but speed and Chinese quality need real testing.
Score
3.8/5
Product/version recorded in source
YuE2
Updated
Oct 27, 2026
Reviewer
OKCodex 每日自动评测系统(公开来源与结构化场景)
Process
Automated evidence review

Best for

Technical teams and independent musicians willing to self-host and build on top

Less suitable for

Casual users wanting plug-and-play results without GPUs or verified Chinese vocals

Generating a full vocal-plus-backing demo from original lyricsZero-shot cover style transfer on an existing songAgentic editing to revise and re-arrange sections of a generated track
Check source ↗

This review uses public sources and structured scenarios. It does not claim hands-on human testing.

At a glance

Product details

Model
YuE2
Category
Creative · Multimodal · Voice
Autonomy
L2
Pricing
Free
Languages
English / Chinese
Product/version recorded in source
Checked
FAQ

Common questions

What pricing information is recorded for YuE?

The catalog lists a free plan. Confirm current prices, included features and usage limits with the provider; this catalog does not verify a total cost or budget ceiling.

What do the catalog and review scores mean for YuE?

The catalog information score is 74/100; it is not measured task performance. The separate source-supported review records 3.8/5 for the product/version recorded in its source: YuE2. This is a source review, not a measured task run. The recorded product/version does not identify an executed model. Manual publication alone does not establish hands-on testing; inspect benchmark runs for actual models, tasks and dates. It is automated source analysis, not a claim of hands-on human testing. The two score scales are not interchangeable.

What product information is available for YuE?

Open-source music large model for full-song generation, from lyrics to complete songs

What does the published review conclude about YuE?

Strong open-source control and value, but speed and Chinese quality need real testing.

Share evidence details

YuE — Automated source review 3.8/5 · Product/version recorded in source: YuE2 Open-source music large model for full-song generation, from lyrics to complete songs This is a source review, not a measured task run. The recorded product/version does not identify an executed model. Manual publication alone does not establish hands-on testing; inspect benchmark runs for actual models, tasks and dates. View evidence on OKCodex: https://www.okcodex.com/en/agents/yue
Share:Weibo
X / Twitter
NL

Weekly OKCodex briefing

Every Wednesday: demand signals, agent reviews, and editorial perspective.

No spam · Unsubscribe anytime

Related agents

Related by catalog categories, models, or text similarity. This does not verify task performance or budget fit.