Experience Shotstack for yourself,
with no risk, and generate your first
video in 5 minutes.
Unlimited developer sandbox
No credit card required
{
"timeline": {
"tracks": [
{
"clips": [
{
"asset": {
"type": "audio",
"src": "https://shotstack-assets.s3-ap-southeast-2.amazonaws.com/music/unminus/berlin.mp3"
},
"start": 0,
"length": 3.96
},
{
"asset": {
"type": "audio",
"src": "https://shotstack-assets.s3-ap-southeast-2.amazonaws.com/music/unminus/berlin.mp3",
"volume": 0.25,
"trim": 3.96
},
"start": 4,
"length": 3.96
},
{
"asset": {
"type": "audio",
"src": "https://shotstack-assets.s3-ap-southeast-2.amazonaws.com/music/unminus/berlin.mp3",
"volume": 0.8,
"trim": 7.96
},
"start": 8,
"length": 3.96
}
]
}
]
},
"output": {
"format": "mp4",
"resolution": "sd"
}
}Raise and lower the volume of a single audio track at different points on the timeline — the track is cut into sections, each with its own level. This is the pattern behind ducking music under a voiceover or fading a bed in and out programmatically.
No account needed until you render.
Render with your AI assistant
Add the Shotstack MCP server to your editor, then paste the prompt
| Output | MP4 |
|---|---|
| Size | 1024 × 576 (16:9) |
| Length | 12s |
| Frame rate | default (25 fps) |
| Tracks / clips | 1 / 3 |
| Assets used | 3 audio |
| Soundtrack | no |
| Merge fields | none (edit the JSON) |
Swap these in the JSON:
To drive it from data, turn any value into a {{ PLACEHOLDER }} merge field.
Save the JSON above as template.json, then render in the free sandbox:
# Save the JSON above as template.json, then:
curl -X POST https://api.shotstack.io/edit/stage/render \
-H "x-api-key: $SHOTSTACK_API_KEY" \
-H "Content-Type: application/json" \
-d @template.json
# Poll until status is "done", then download response.url
curl https://api.shotstack.io/edit/stage/render/RENDER_ID -H "x-api-key: $SHOTSTACK_API_KEY"import { readFileSync } from 'node:fs';
const headers = { 'x-api-key': process.env.SHOTSTACK_API_KEY, 'Content-Type': 'application/json' };
const edit = JSON.parse(readFileSync('template.json', 'utf8'));
const { response } = await fetch('https://api.shotstack.io/edit/stage/render', {
method: 'POST', headers, body: JSON.stringify(edit)
}).then((r) => r.json());
let render;
do {
await new Promise((r) => setTimeout(r, 3000));
render = (await fetch(`https://api.shotstack.io/edit/stage/render/${response.id}`, { headers }).then((r) => r.json())).response;
} while (!['done', 'failed'].includes(render.status));
console.log(render.status, render.url);import json, os, time, requests
headers = {"x-api-key": os.environ["SHOTSTACK_API_KEY"]}
with open("template.json") as f:
edit = json.load(f)
render_id = requests.post("https://api.shotstack.io/edit/stage/render", json=edit, headers=headers).json()["response"]["id"]
while True:
render = requests.get(f"https://api.shotstack.io/edit/stage/render/{render_id}", headers=headers).json()["response"]
if render["status"] in ("done", "failed"):
break
time.sleep(3)
print(render["status"], render.get("url"))npm install -g @shotstack/cli
export SHOTSTACK_API_KEY=your_sandbox_key
shotstack validate template.json # offline schema check, free
shotstack render template.json --env stage --watchSandbox renders are free and watermarked. Production: the v1 endpoint and key. SDKs
One render is a demo. A audio for every row of your data is the workflow.
Paste into Claude, ChatGPT or Cursor — your AI assistant builds the workflow around your data.
Node.js, one render per CSV row, wired to this template. Setup steps in the header.
Guides: Bulk-create from a CSV · Webhooks · Templates endpoint · Render your first video
AI assistants: Markdown () · · · CLI skill: SKILL.md () · MCP server
Replace the asset URLs in the JSON with your own files and adjust text, timing or effects on each clip. To generate many variations from data, turn the values you change into merge fields ({{ PLACEHOLDER }}) and pass a merge array with each request.
Yes. This template renders 1024×576 by default. Set output.resolution to "1080" or output.size to a larger 16:9 size in the JSON; the layout scales with it.
This template renders a 1024×576 (16:9) MP4 that is 12 seconds long, built from 3 clips on 1 track.
Yes. Loop over your data source (a CSV, a database, a webhook payload), send one render request per row with different values, and use a callback URL so Shotstack notifies you when each audio is ready.
Unlimited developer sandbox
No credit card required
