Give Reelspan a long video (a podcast, a talk, a stream, a tutorial) and it finds the 3 to 5 moments that work on their own as short clips for TikTok, Reels and Shorts. Each clip is cut to 9:16 and kept in your media library with its own media_id, so you post it like any other video.
How it works:
- The speech is transcribed with the time each sentence is said.
- The AI picks the moments with a strong start and one complete idea, 15 to 60 seconds each, and writes a title for each one.
- Each moment is cut from the middle of the picture to 9:16 (at most 1080 pixels wide), with its sound. Nothing is written over the picture.
A video with no speech still works when the AI model can see pictures: the moments are then picked from frames of the video.
Limits and cost
| Video length | 30 seconds to 60 minutes |
| Clips | 3 to 5, each 15 to 60 seconds |
| Cost | 1 AI credit per started 5 minutes of the video: a 12-minute video uses 3 credits, an hour uses 12 |
| At once | 2 videos per workspace |
The credits are checked when you start and charged when the clips are ready. A job that fails costs nothing. Without enough credits left this month the call answers 429 with error_code: AI_CREDITS_EXHAUSTED and credits_needed.
Start a job
Send the media_id of a video in your library, or of one you just uploaded with Upload media. The answer comes at once with a job_id; the clips take a minute or two for a short video and longer for an hour-long one.
curl -X POST https://api.reelspan.com/v1/clips \
-H "Authorization: Bearer $REELSPAN_API_KEY" \
-H "Content-Type: application/json" \
-d '{"media_id": "med_4f2a…"}'const job = await reelspan.request('POST', '/v1/clips', { json: { media_id: 'med_4f2a…' } });job = reelspan.request("POST", "/v1/clips", json={"media_id": "med_4f2a…"}){
"success": true,
"job_id": "clp_9Qm2…",
"media_id": "med_4f2a…",
"status": "queued",
"stage": null,
"progress": 0,
"source_seconds": 1480,
"credits": 5,
"clips": [],
"error": null,
"error_code": null,
"created_at": "2026-11-17T10:02:11.000Z",
"finished_at": null
}Check on it
Poll GET /v1/clips/{job_id} every few seconds. status goes from queued to running to completed or failed; while it runs, stage says what it is doing (reading, transcribing, choosing, cutting, saving) and progress goes from 0 to 100.
curl https://api.reelspan.com/v1/clips/clp_9Qm2… \
-H "Authorization: Bearer $REELSPAN_API_KEY"{
"success": true,
"job_id": "clp_9Qm2…",
"status": "completed",
"progress": 100,
"credits": 5,
"clips": [
{
"media_id": "med_c81d…",
"title": "The one pricing mistake every new shop makes",
"why": "Opens with a clear problem and ends on the fix.",
"start_seconds": 312.4,
"end_seconds": 356.0,
"duration_seconds": 43.6,
"url": "https://…"
}
]
}Post a clip by its media_id, like any upload:
curl -X POST https://api.reelspan.com/v1/posts/video \
-H "Authorization: Bearer $REELSPAN_API_KEY" \
-F "user=my-brand" -F "platform[]=tiktok" -F "platform[]=instagram" \
-F "video_media_id=med_c81d…" -F "title=The one pricing mistake every new shop makes"GET /v1/clips lists the latest jobs, newest first; add media_id to see only the jobs for one video.
Errors
| Status | error_code |
What happened |
|---|---|---|
| 400 | VIDEO_TOO_SHORT, VIDEO_TOO_LONG |
The video is shorter than 30 seconds or longer than 60 minutes |
| 400 | VIDEO_UNREADABLE |
The file can't be read; export it again as MP4 |
| 404 | MEDIA_NOT_FOUND |
No video with that media_id in your workspace |
| 429 | AI_CREDITS_EXHAUSTED |
Not enough AI credits left this month |
| 429 | CLIPS_BUSY |
Two videos are already being clipped; wait for one to finish |
| 503 | AI_NOT_CONFIGURED |
AI is not set up on this server |
A job that fails while running ends with status: failed, a plain error and an error_code, such as NO_MOMENTS (no part of the video works as a clip on its own), NO_SPEECH or LIBRARY_FULL.