ACE Music MCP
通过 MCP 在 AI Agent 中使用灵感或自定义歌词生成双曲候选、生成歌词并查询任务。
Describe the music you want to hear, or bring lyrics you wrote yourself. ACE Music generates two candidates for you to preview, compare, and download your favorite version.
Get API credentialsOpen the creation workspaceInspiration descriptions and custom lyrics cover the basic workflow from concept to finished song.
Each complete generation delivers a set of candidates to preview individually and download as MP3s.
Song, lyrics, and task query APIs, all using the same platform credentials.
Describe the theme, style, and mood—ideal for quickly exploring song directions.
Enter a title, lyrics, and style while retaining control over the content.
Generate a lyric draft first, then submit it for song generation once you are satisfied with your edits.
Save the task ID after asynchronous submission, then use free queries to retrieve progress and results. Continue querying the original task after a network interruption to avoid duplicate creations.
curl --request POST 'https://api.acedata.cloud/acemusic/audios' \
--header 'Authorization: Bearer YOUR_API_KEY' \
--header 'Content-Type: application/json' \
--data '{
"model": "acemusic-v1.1",
"custom": false,
"prompt": "一首关于故乡的温暖民谣",
"async": true
}'Read the full integration guide →Compare different interpretations and choose a direction for further creation.
Use themes and moods to explore music ideas for videos, stories, and game scenes.
Integrate through the API or Agent to connect lyric writing, generation, and previewing into your own workflow.
Activate the service and choose the credit plan that suits you.
Choose a model and submit your inspiration or custom lyrics.
Query the original task, compare completed candidates, and download the audio.
| Method | You provide | Best for |
|---|---|---|
| Inspiration mode | A creative description | Quickly exploring themes and styles |
| Custom mode | Title, lyrics, and music style | When you already have clear lyrics and a content direction |
| Lyrics assistant | Theme and optional style | Refining the text first, then creating music |
Default creation model. Public API ID: acemusic-v1.1.
Retained optional model. Public API ID: acemusic-v1.
Music style and performance are generated by the model; actual duration is determined by the result.
At the highest plan rate, each set of two complete candidates is charged once.
At the highest plan rate, each successful lyric generation is charged separately.
Valid credentials and task permissions are required; you can view your own results even with a zero balance.
Pricing benchmark updated on 2026-11-17; actual pricing is affected by the plan used, discounts, and site pricing. No generation fee is charged for final generation failures or partial failures, and charged tasks are not automatically reissued.
Acceptance returns a task ID. Playback is available only after generation and audio delivery are complete. Please query the original task; do not treat acceptance as completion.
Yes. Use custom mode to submit a title, lyrics, and style; drafts provided by the lyrics assistant can also be edited first.
The successful playable candidate is retained, the final task is marked as partially failed, and no generation fee is charged for that set.
The current basic creation output is song audio, with MP3 download supported. Uploading, covers, continuation, stem separation, and voice capabilities are not included in this release.
Please determine this based on the applicable terms of service and project context, and ensure that you have the necessary rights to the input content and intended use.
Create directly with Studio, or integrate ACE Music into your application.
Get startedView API documentation