FUZZ-2.0 is a music generation model in Riffusion's FUZZ series, suitable for turning song themes, style descriptions, or existing lyrics into listenable music. On this platform, it is used through the Producer music entry point, supporting both prompt-based creation and custom lyrics workflows. It is a general-purpose choice for song drafts, soundtrack exploration, and everyday music solution comparisons; for professional production or custom post-processing, consider Pro or Raw.
Submit requests to the public API at api.acedata.cloud using the documented parameters, then use the results in your application.
Input parameters and result formats vary by service. Use the public API for this model and follow its guide for generation, task retrieval and editing operations.
Clarify capacity, inputs and outputs, and invocation methods before selecting a model.
The above input, control, and delivery specifications correspond to this platform's usage entry point and do not treat shared editing features as all of FUZZ-2.0's native capabilities.
Discover what FUZZ-2.0 can bring to your work.
Organize prompts around themes, moods, and arrangement intentions—for example, describing a rainy-night atmosphere, a jazz style, and a combination of piano and saxophone—then generate music drafts for listening. It is suited to turning abstract ideas into sound concepts, helping creators compare different directions rather than only receiving lyrics or text suggestions.
When you already have lyrics, you can use custom mode to submit the lyrics and title, and use labels such as Verse and Chorus to express section intentions. This enables song creation around your own text, making it easier to preview how verses and choruses are presented; labels are used to express structure and do not mean melodies are fixed beat by beat or vocal arrangements are locked in.
Music generation is delivered as task results, and audio links can be saved together with lyrics, titles, and status information. Applications can use asynchronous mode to obtain a task ID and then query the completed result, or receive results through callbacks; this is suitable for building creation records and listening lists without requiring users to remain on a waiting page.
Start with specific tasks to find where the model can be effective.
Enter lyrics that have already been written, a song title, and style intentions to generate song demos for discussion. Lyricists can use them to review section organization and song direction, while production teams can communicate around the same text. Deliverables include listenable audio and corresponding lyric records; musical review is still required before formal production.
Organize shot moods, narrative themes, and desired instruments into text, and explore scoring directions using instrumental music settings. This is suitable for comparing different options—such as warm, tense, or upbeat—before editing, then importing audio into editing tools for test matching; visual rhythm alignment and final duration trimming should be completed in post-production.
Write a set of different prompts around holidays, events, or content series to generate and save theme music candidates. Applications can associate task IDs, prompts, work titles, and audio links to create a replayable concept library, making it easier to compare creative directions and retain text records of each attempt.
Choose based on task complexity, input materials, and expected results.
FUZZ-2.0 is suitable for prompt exploration, custom lyric demos, and standard music candidate generation. Compared with Pro and Raw in the same series, its focus is on the general creative workflow: first verify whether the theme, lyrics, and style are suitable, then decide whether to move into professional production or post-processing. If the project has no specific special delivery requirements, you can start by previewing this version.
FUZZ-2.0 Pro is designed for professional production needs, while FUZZ-2.0 Raw is for creators who want to handle post-processing themselves; they are different versions and their names should not be used interchangeably. When choosing, it is recommended to compare outputs using the same lyrics and style intent, then decide based on your own workflow. Do not judge audio quality, generation speed, or editing capabilities solely by version numbers.
Prepare the inputs first, then proceed to the appropriate application workflow.
Prepare the content theme, target mood, lyrics or instrumental music requirements, and describe the intended use and desired atmosphere.
Submit the creation request with the exact FUZZ-2.0 invocation value; select the prompt, lyric, and title fields for the appropriate entry point, and save the task_id and work ID. Complete the preview first, then select subsequent actions according to this model's documentation, without mixing in editing fields from other tiers.
Design the task directly based on the inputs and acceptance priorities below.
Submit model=FUZZ-2.0 through /producer/audios; use prompt to describe the musical direction, and submit lyric and title separately in custom mode.
Preview whether the overall musical direction matches the content, and check the song status, actual duration, and playable audio; select the draft first, then decide on subsequent processing.
Before formal use, understand the output quality and capability scope.
Answers to common questions when using FUZZ-2.0.
No. FUZZ-2.0 is a general-purpose creative option, Pro is designed for professional production needs, and Raw is intended for customized post-production. Use the exact model name when calling it; do not treat the work duration or effects in Pro examples as fixed FUZZ-2.0 performance.
Yes. You can submit text through the custom lyrics workflow, set custom=true, and include a song title and creative prompt. Lyrics can use labels such as Verse and Chorus to organize sections. It is recommended to generate a sample first, listen to how the lyrics and sections are rendered, then adjust the text or stylistic intent.
The creation entry provides an instrumental toggle, which can be used to express the generation intent for lyric-free music, with prompts describing mood, style, and instrumental arrangement. Instrumental tasks still require relevant fields to be submitted as requested; after completion, listen to confirm that the result meets the intended soundtrack use.
The API delivers JSON. Completed results include work information such as audio_url, and music can be obtained through the audio link. When using async=true, you first receive task_id and then query the result; you can also set callback_url to receive the result associated with the task after it is completed.
Use POST /producer/audios, set model to FUZZ-2.0, use action=generate for a new work, and prepare prompt and lyric. The request requires both text fields; when using custom lyrics, also set custom=true, and submit the request using valid authorization credentials.
Choose a development approach or application suited to this model and follow its setup guide.
↗Python SDK Integration Guide
SDK installation, authentication and examples↗TypeScript SDK Integration Guide
SDK installation, authentication and examples
↗This is only a basic call example. See the full docs for more parameters and advanced usage.
View Docs