submitted or queued state. Set wait_audio: true to block until audio is fully ready (up to 100 seconds).
The server reads your Suno session from the
SUNO_COOKIE environment variable. Pass a Cookie header containing a valid __client value to use a different Suno account on a per-request basis.Request body
A text description of the music to generate. Suno uses this as a
gpt_description_prompt and will write lyrics automatically.When
true, generates a song without vocals.The Suno model version to use for generation (e.g.
chirp-v3-5, chirp-v4).When
true, the request polls Suno until both clips reach streaming or complete status before returning. The poll window is up to 100 seconds. When false, clips are returned immediately in submitted / queued state and you must poll GET /api/get for the final URLs.Response
An array of twoAudioInfo objects.
Unique identifier for the generated audio clip.
Title assigned to the clip by Suno.
URL of the cover art image for the clip.
The lyrics used in the generated audio.
URL of the generated audio file. May be empty until
status is streaming or complete.URL of the generated video file. May be empty until generation is complete.
ISO 8601 timestamp of when the clip was created.
The model used to generate this clip (e.g.
chirp-v3-5).The original text prompt submitted to Suno’s description system.
The actual lyrics prompt used during generation (may differ from the input
prompt after GPT expansion).Current generation status. One of:
submitted, queued, streaming, complete, error.Internal clip type as reported by Suno.
Music style tags associated with the clip.
Music style tags that were excluded during generation.
Duration of the audio clip in seconds.
Human-readable error description. Present only when
status is error.Error responses
| Status | Meaning |
|---|---|
400 | Bad request (e.g. missing required field) |
402 | Out of Suno credits — payment required |
500 | Internal server error |
503 | Network error connecting to Suno |
