SeedAudio API
Async generation · Your site API keyBring dialogue, music, ambience and sound effects into your product with the SeedAudio API. Turn a clear creative brief into an audio task you can track and download.
A paid credit pack or an active paid subscription unlocks API access. Use the same credit balance as the website; gift credits alone do not unlock access.
PROMPT → TASK → AUDIO
curl --request POST 'https://seedaudio.now/api/v1/audio/generations' \
--header "Authorization: Bearer $SEEDAUDIO_API_KEY" \
--header 'Content-Type: application/json' \
--header 'Idempotency-Key: first-audio-scene-001' \
--data '{
"model": "seed-audio-1.0",
"prompt": "On a train on a rainy night, a passenger softly says: The next stop is ours. Wheels and rain stay in the background. End with a brief tail of carriage ambience. No background music."
}'Set SEEDAUDIO_API_KEY to your site API key first. Use a new Idempotency-Key for each new task; keep the same key and content when retrying a request.
01 / INTEGRATION DIRECTION
Organize inputs around a sound scene and delivery around a task. Submit with your site API key, then use one task ID to check status and retrieve audio.
Bring dialogue, emotion, music, ambience and action sounds together in one creative brief.
Use authorized audio or image references when a character or scene needs a specific direction.
Submit, wait and retrieve the audio through a background task that fits your editing and content pipeline.
02 / CREATIVE CONTROLS
Shape creative direction with text and references, then choose output settings for delivery. See the API docs for complete fields, limits and examples.
Write up to 2,048 Unicode characters. Define characters, language, emotion, timing and surroundings so every sound belongs to the same scene.
Use up to 3 audio references; a selected voice takes 1 slot. Alternatively, provide 1 image. An image cannot be combined with a voice or audio references.
Describe vocal character, emotion and natural pauses. Set speed and volume from 0.5–2, and pitch from −12 to 12.
Choose mp3, wav, pcm or ogg_opus. Sample rates of 8, 16, 24, 32, 44.1 or 48 kHz support different post-production needs.
Keep the returned task ID and check queued, processing, succeeded or failed status. When audio is ready, download it through an authenticated site URL.
Include an Idempotency-Key with every submission. The same key and content return the original task; changing content under the same key is rejected.
03 / BUILT AROUND YOUR CONTENT
Plan creation around real content. Explore our audio examples and prompt guide to evaluate the scenes you have in mind.
Dialogue, ambience and action sounds that follow the story on screen.
Design character exchanges, intros and narrative moments.
Explore the sound of characters, spaces and changing events.
Express a score through atmosphere, instruments and rhythm.
Describe actions, materials, distance and when each sound occurs.
Keep creative briefs and their resulting audio assets connected.
04 / INTEGRATION FLOW
Submit, query and download through the site API. Keep your key on your server and use references you have permission to use.
Buy a credit pack or start a paid subscription, create your site API key, and prepare a prompt and any HTTPS references. Keep at least 150 available credits before submitting.
Send the seed-audio-1.0 model and your creative brief with a Bearer key and Idempotency-Key. Save the returned task ID.
Query the task until it finishes, check the final credit charge, then use your site key to download the audio from its task URL.
05 / ACCESS & PRICING
A paid credit pack or an active paid subscription unlocks API access. Each task reserves 150 credits. On success, actual audio duration is billed at 75 credits per minute, rounded up, with a minimum of 10 and a maximum of 150 credits.
06 / QUESTIONS
NEXT SCENE
Listen to examples, write your scene, then bring the sound into your product.