ShortsAIAutomate shorts end to end: one request in, one webhook out
The whole automation loop is two HTTP messages. You POST a video URL with an API key and a webhook address; when processing ends, ShortsAI sends exactly one signed webhook carrying clip titles and durable download links. No polling loop, no timeout guessing.
API calls draw from the same minute balance as the dashboard, with no separate meter and no per-call pricing. On the self-hosted edition there is no meter at all, which is what makes an always-on pipeline affordable.
For agent-driven automation (Claude, ChatGPT, custom agents) the same account also exposes an MCP server; that protocol surface is documented on its own page.
The loop, end to end
Start a job with one request:
curl -X POST https://api.shortsai.com/api/process \
-H "Authorization: Bearer osk_..." -H "Content-Type: application/json" \
-d '{"url": "https://youtube.com/watch?v=...", "acknowledged": true,
"webhook_url": "https://your-server.com/hooks/shortsai",
"webhook_secret": "your-shared-secret"}'The response returns a job id immediately. Minutes later, when the clips are cut, subtitled and archived, ShortsAI POSTs once to your webhook URL with the job outcome, clip titles and download links durable enough to fetch later. A failed job also fires the webhook, so your pipeline never hangs on silence.
Verifying the webhook
If you passed a webhook_secret, the request carries an
X-ShortsAI-Signature header of the form
sha256=<hex>: the HMAC-SHA256 of the raw request body under
your secret. Recompute it and compare in constant time:
expected = "sha256=" + hmac.new(secret, raw_body, hashlib.sha256).hexdigest()
hmac.compare_digest(expected, request.headers["X-ShortsAI-Signature"])Reject anything that does not match and you have closed the door on forged deliveries.
Using it from n8n, Zapier or Make
No dedicated node is needed: the flow is a generic HTTP Request step that
POSTs to /api/process, then a Webhook trigger that receives the
completion payload and fans out to whatever comes next, posting to socials,
dropping links in Slack, logging to a sheet. The two-step shape above is the whole integration, which is why generic
nodes cover it. There is now also an importable
n8n workflow that wires the whole
loop, including channel watching, Telegram approval and scheduled publishing.
A weekly pipeline in one cron line
Because the API is one POST, scheduling is whatever scheduler you already have. A cron job that submits the latest episode URL every Monday, a GitHub Action on your podcast repo, or an agent that watches a feed. The webhook does the second half, so nothing stays running in between. If you would rather not write the curl, the CLI wraps it:
uvx shortsai process "$EPISODE_URL" \
--webhook https://your-server.com/hooks/shortsai \
--webhook-secret "$SECRET"What automation costs
API and MCP calls draw from the same minute balance as the dashboard. There is no per-call price, no separate API tier and no automation surcharge. As of August 2026 that is not the market default: the mainstream tools meter their APIs per source minute or per operation, on top of subscription tiers, so an always-on pipeline runs with a taxi meter attached. Self-hosted ShortsAI has no meter of any kind, and the hosted plans are flat:
ShortsAI comes in two editions and they are priced very differently, so it is worth being precise. ShortsAI (self-hosted) is free and open source under the MIT licence: Free and open source. Run it with Docker on your own machine and bring your own API keys. No watermark, no usage cap, no subscription. ShortsAI Cloud is the hosted service: High-speed GPU-accelerated video clipping in seconds. Free plan includes starter minutes. Paid plans start at $12/month for high-volume exports without watermarks.
Agents instead of scripts
If the thing driving the pipeline is an AI agent rather than a script, the same account exposes a native MCP server with six tools covering process, status, clips, quota, subtitles and publishing. The endpoint, the tool table and client setup live on the MCP server and API page.
Common questions
- Can I automate YouTube Shorts creation with an open source tool?
- Yes. ShortsAI is MIT-licensed and its self-hosted edition serves the same REST API and MCP server as the hosted service, with no metering. One POST submits a video, a signed webhook returns the finished clips, and the publishing endpoint posts them to YouTube Shorts, TikTok and Instagram Reels.
- Is there an n8n integration for ShortsAI?
- Yes: an importable workflow at /n8n-youtube-shorts-automation watches a YouTube channel, clips each video, sends the clips to Telegram for approval and schedules the approved ones to your socials. Nothing forces you to use it, though, since the integration is just an HTTP Request node posting to /api/process plus a Webhook trigger.
- Do automated API calls cost more than using the dashboard?
- No. API and MCP usage draws from the same minute balance as the dashboard with no per-call pricing: 20 free minutes a month on the hosted free tier, flat paid plans from $12/month, and no meter at all on the self-hosted edition.
Sources
- Webhook signing implementation in the project source at github.com/shortsai/shortsai.
- Competitor API metering checked 2026-08-04 on public pricing and developer documentation.