AnyInt Docs

Media & Music

Use this section for asset-generation workflows that do not behave like ordinary chat completions, especially AI Music tasks.

Use this section for asset-generation workflows that do not behave like ordinary chat completions, especially AI Music tasks.

Start here

TaskPageOutcome
Understand when to use music-specific workflowsOverviewDecide whether your workload is chat, media generation, or AI Music
Create songs, lyrics, styles, or cover artMusic GenerationSubmit generation requests with the right prompt, model, and callback choices
Transform existing audio, create personas, or export WAV/MVMusic TransformReuse existing audio or completed tracks in follow-up workflows
Poll tasks and process callbacksMusic Tasks and WebhooksTrack async jobs until final assets are ready
Review common workflow patternsMusic Use CasesChoose a route sequence for common product experiences

API contracts

Use AI Music API for exact Suno-compatible route contracts. Use Media API for DashScope image and generic video routes.

How media and music differ from chat

DifferenceWhat to design for
Async task lifecycleStore task IDs and poll or process callbacks until success or failure
Large generated assetsStore output URLs and metadata separately from chat messages
User-visible wait timeShow pending, processing, succeeded, and failed states in your UI
RetriesMake retry behavior idempotent enough to avoid duplicate paid jobs when possible
Rights and reviewDecide how your product handles uploaded audio, generated tracks, and user-facing moderation
  1. Create the task with a safe placeholder callback URL or a production HTTPS callback endpoint.
  2. Store the returned task ID with the user, project, prompt, and selected model.
  3. Poll the documented query route or accept webhook callbacks.
  4. Mark the asset ready only after the task status is successful and output URLs are present.
  5. Keep the original request payload and final metadata for customer support and retry analysis, without logging private API keys.

Choose the right API family

OutputStart with
Text-only lyrics, descriptions, or planning promptsOpenAI-compatible or another text API family
Image generation or prompt-driven videoMedia API and Video Generation
Full song, cover, stem, persona, WAV, or music video workflowAI Music API
Mixed product workflowsKeep text, media, and music tasks as separate route calls with separate error handling

Production checks

Before launch, verify task creation, polling, callback reachability, duplicate callback handling, failed-task UI, output URL storage, and user-facing error messages for quota or policy failures.

On this page