AI music prompt max length: 5000 characters on Sume
Sume's Music Router accepts a prompt of 1-5000 characters. What to spend them on: a seven-axis brief, section markers and exclusions, in one positive prompt.

The Sume Music Router takes a prompt of 1 to 5000 characters. That is the only numeric limit on the text: there is no separate duration, negative-prompt or seed field to spend it on, so the prompt carries the whole brief.
Limits are from the Music Router docs, read 2026-09-30. Do not compare it with another vendor's number without checking what that number measures.
How does this compare with ElevenLabs' 6,132?
The ElevenLabs changelog for Sep 14, 2026 says composition plan generation chunks support up to 6,132 characters, with up to 30 lines of 200 characters each. That limit applies to a plan chunk in a different request shape, so it is not a prompt limit equivalent. See composition plan vs prompt only.
What fields count and what is rejected?
| Field | Rule |
|---|---|
prompt | Required, 1-5000 characters |
negative_prompt | Unsupported when non-empty; put exclusions in the positive prompt |
duration / duration_seconds | Rejected; steer length in the prompt |
image_url | Optional public HTTPS image, or null to clear |
What should I put in the prompt?
Music 1.0 accepts a text prompt and optional image conditioning, and the docs recommend a scene-specific brief across seven axes: emotion, genre, tempo as a number, key and mode, two to four instruments with texture, an arc with one named moment, and era or production. These are creative directions, not guaranteed settings.
For structure, use section markers such as [0:00-0:30] Intro: .... Close with one clause such as "Instrumental, no vocals." A brief with all seven axes is far shorter than 5000 characters, so the limit rarely binds.
What happens past the limit?
The schema bounds the prompt at 5000 characters, so shorten it rather than relying on truncation. Each generation is charged the fixed Music price, $0.125 per audio, regardless of prompt length.
Sources
Related posts
More in Developers
- n8n Wait node under 65 seconds: how to poll a Sume job
n8n keeps waits under 65 seconds in memory and saves longer ones to the database. Poll a Sume job with next_poll_after_seconds, or switch to a webhook resume.
- Next.js image SSRF fix: narrow remotePatterns for Sume
Next.js patched an Image Optimization SSRF via allow-listed remote URLs. Allow only media.sume.com for Sume images, and never raw provider hosts.
- 429 backoff with jitter in Python: OpenAI's advice, Sume's headers
OpenAI recommends exponential backoff with random jitter on 429. A Python status poller that honors Sume's retry-after first and falls back to backoff.
- OpenAI 429 slow_down vs 503 server_is_overloaded, and Sume
OpenAI now splits 429 slow_down (traffic ramping too fast) from 503 server_is_overloaded. Here is how each maps to Sume's 429 scope and 503 retry rules.
Written by Sume