Sora extensions to 120 seconds: 30-second clips and Timeline
Sora's Videos API let you extend a clip up to 6 times, 120 seconds in all, and shut down 2026-09-24. On Sume, chain 30-second clips and join them in Timeline.

Sume has no extend endpoint. To replace a Sora extension chain, make one seedance-2.5 or wan-3.0 clip of up to 30 seconds per request, start each next clip from a still of the previous one, and join the clips with Timeline 1.0. Four 30-second seedance-2.5 clips at 720p make a two-minute video.
OpenAI's guide described extensions that continue a completed video up to 6 times, 120 seconds at most, and its deprecations page lists the Videos API and the sora-2 models with a shutdown on 2026-09-24. Both pages were read 2026-10-01.
What replaces a Sora extension on Sume?
Each Sume video job is a new request. The video docs list seedance-2.5 at 4 to 30 seconds and wan-3.0 at 2 to 30 seconds; every other catalog model tops out at 15. Pass the last frame of the finished clip as a first_frame entry in frame_images so the next clip starts where the last one ended; the walkthrough is in extend an AI video past 30 seconds.
How do I join the clips?
Timeline 1.0 takes Sume-hosted media.sume.com media, so import the generated files first with POST /v1/media-imports. It requires an Idempotency-Key, and it needs an audio spine: send audio.duration_seconds plus a voice or music file, or set audio.mode to silence. Render pricing is $0.10 per output minute, rounded up. See the Timeline docs.
What does 120 seconds cost?
Reserved amounts come from Sume's pricing code at list x 1.25, 16:9, 720p. A failed job releases its reservation.
| Item | Amount |
|---|---|
| One 30 s clip, seedance-2.5, 720p | $17.34 |
| Four clips | $69.36 |
| Timeline render, 2 output minutes | $0.20 |
| Total reserved | $69.56 |
What do I give up compared with extensions?
A Sora extension continued the same generation. A Sume chain is a series of independent clips, so check each seam for drift in lighting and faces, and write each prompt to repeat the details that must stay fixed. Use wan-3.0 or seedance-2.5 when you need the 30-second length; a single request on the other models stops at 15.
Sources
Related posts
More in Developers
- Sume editing tool source limits: 1,800 s and 300 s
Video trim, audio detach and video inspect accept a source up to 1,800 seconds. Video filter, Timeline compose and video frames stop at 300 seconds.
- Spotify HDR video podcast: tone-mapped to SDR Rec.709
Spotify converts HDR (Rec.2100) uploads to SDR Rec.709. A Sume exact trim re-encodes to yuv420p and has no colour fields, so tone-map before you trim.
- Spotify video audio channels: mono upmixed, 5.1 downmixed
Spotify upmixes mono to stereo, downmixes 5.1, and rejects multi-channel. Sume audio detach offers only channels source or mono; it cannot mix to stereo.
- Sume SDK fetch wrapper: do not add an Authorization header
A fetch wrapper passed to createSumeClient must not add Authorization. Sending it with x-api-key returns 401, because Sume accepts only one credential.
Written by Sume