FUZZ-1.1 Pro is the legacy professional tier in the Riffusion FUZZ series, suitable for organizing song themes, lyrics, and sonic direction into listenable musical works. On this platform, you can create through the Producer or Riffusion audio entry points and obtain audio links and song information. It is especially suitable for projects that continue to use this version, rather than the default choice that bundles all new-generation music tools together.
Submit requests to the public API at api.acedata.cloud using the documented parameters, then use the results in your application.
api.acedata.cloudFUZZ-1.1 ProInput 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 choosing a model.
The above describes the inputs, outputs, and controls of this platform's creation entry points and does not represent native capacity metrics; editing actions are performed according to the actual support of the entry point and model used.
Learn what FUZZ-1.1 Pro can bring to your work.
When lyrics are not yet written, use prompts to describe the song's theme, mood, and sonic direction; when lyrics already exist, submit the text and title through custom mode. These two approaches allow FUZZ-1.1 Pro to participate in inspiration exploration and to turn a defined lyric concept into a preview track.
Lyric content and sound descriptions can be prepared separately, making it easier to distinguish between “what to sing” and “what it should sound like.” The Riffusion entry provides style, while the Producer entry uses prompt to describe the creative direction. This is suitable for finalizing lyrics first and then comparing different arrangement and mood options, rather than rewriting the full creative brief every time.
Generated results provide not only audio links, but also song ID, title, lyrics, status, and actual duration, making it easy to save works and link follow-up tasks. The asynchronous approach can first obtain task_id, then query or receive callbacks, allowing previewing, selection, and archiving to fit into the normal workflow of music creation applications.
Start from specific tasks and find where the model can make an impact.
Enter completed verse and chorus lyrics, then add a song title and sonic direction to generate preview demos for creators to discuss. The focus is on judging the overall feel after the words are set to music, then deciding whether to revise the lyrics or try another arrangement direction; the deliverables are audio works and their associated information.
Prepare theme and mood descriptions for short videos, podcasts, or event content, and use lyric-free mode to explore soundtrack directions. After previewing, provide suitable work links and actual durations to editors for selecting material and arranging pacing; do not treat duration requirements in natural language as precise editing instructions.
For song projects that already use FUZZ-1.1 Pro, retain the model selection, lyrics, and sound description to continue producing candidate works. When an existing song needs editing, prepare audio_id and clear modification goals, then use an appropriate continuation or segment replacement task, saving the new results for comparison and previewing.
Choose based on task complexity, input materials, and expected results.
If a project has already established listening preferences around FUZZ-1.1 Pro, or needs to maintain the same version's creative conditions as existing works, it is worth keeping. FUZZ-1.1 is the standard tier of the same generation, while FUZZ-1.0 Pro is another historical version; do not judge differences solely by the Pro name, and compare results using the same lyrics and vocal requirements.
For brand-new projects, you can compare FUZZ-2.0, FUZZ-2.0 Pro, and this model by listening: the former is the general recommended tier, the latter is for professional production, while FUZZ-1.1 Pro retains the value of a historical professional-tier option. Make trade-offs based on lyric presentation, musical direction, and required editing tasks, and do not apply all features of later versions to this model.
Arrange the inputs first, then proceed to the corresponding application workflow.
Prepare lyrics, vocal direction, and the existing work ID, and clarify whether this time you are changing only the style or one section.
Submit the creation using the exact FUZZ-1.1 Pro invocation value; select the prompt, lyric, and title fields for the corresponding entry point, and save the task_id and work ID. Complete the listening review first, then choose follow-up actions according to this model's documentation, without mixing in editing fields from other tiers.
Design tasks directly from the following inputs and acceptance priorities.
First generate a clear version through the corresponding entry point, then select editing operations supported by this model; modify only one creative goal at a time.
Compare the lyrics, melody, and section transitions before and after modification, and keep a record of changes; Pro in the title does not mean that all tools from other versions can be used.
Before formal use, understand the output quality and capability scope.
Answers to common questions about using FUZZ-1.1 Pro.
No. They are different tiers within the same generation. FUZZ-1.1 Pro is the historical professional tier, while FUZZ-1.1 is the standard tier. Clearly specify FUZZ-1.1 Pro when calling it; when choosing, you can compare results using the same lyrics and sound direction without assuming a fixed degree of audio quality improvement.
Use generate and enable custom mode, then submit lyric; you can also provide a title and sound direction. The Producer endpoint requires action, lyric, and prompt; the Riffusion endpoint requires action and lyric, and can use style to describe the style.
You can use instrumental mode without lyrics and describe the music direction around mood, instrumentation, and purpose. It is suitable for soundtrack candidates and instrumental previews; requests must still retain the fields required by the endpoint, and the final length should be read from the generated result rather than treating length descriptions in the prompt as a hard guarantee.
Existing Riffusion creation workflows can continue using /riffusion/audios; use /producer/audios for the Producer workflow. Both can specify this model, but their field and action names differ, so keep request construction and result parsing consistent with the selected endpoint.
After the task is complete, read audio_url from data to obtain the audio, and save the song ID, title, and status. When using async=true, save task_id first, then query for the completed result; you can also configure callback_url to receive callbacks and associate the returned work with the original creation task.
Choose a development approach or application suited to this model and follow its setup guide.
Python SDK Integration Guide
SDK installation, authentication and examplesTypeScript SDK Integration Guide
SDK installation, authentication and examples