Sora 2 snapshot ids removed Sept 24, 2026: the full OpenAI list
OpenAI's deprecations page lists five Sora ids removed from the API on Sept 24, 2026, with no replacement named. What to do with old ids and Sume's video list.

OpenAI's deprecations page lists five Sora ids for removal from the API on September 24, 2026: sora-2, sora-2-pro, sora-2-2025-10-06, sora-2-2025-12-08 and sora-2-pro-2025-10-06. Today is October 2, so any call that still names one of them is past the removal date. OpenAI's page names no replacement, and Sume's video docs list no Sora model.
The ids on OpenAI's page
The OpenAI deprecations page says that on March 24, 2026 it notified developers using the Videos API and the Sora 2 video generation model aliases and snapshots of their deprecation and removal on September 24, 2026.
| Model id | Kind | Removal date | Replacement listed |
|---|---|---|---|
| sora-2 | Alias | 2026-09-24 | None listed |
| sora-2-pro | Alias | 2026-09-24 | None listed |
| sora-2-2025-10-06 | Snapshot | 2026-09-24 | None listed |
| sora-2-2025-12-08 | Snapshot | 2026-09-24 | None listed |
| sora-2-pro-2025-10-06 | Snapshot | 2026-09-24 | None listed |
Pinning a snapshot did not help
Teams often pin a dated snapshot to avoid surprise changes. Here the dated snapshots and the aliases are on the same list with the same date, so pinning bought no extra time. If a config file, a Make scenario or an environment variable still holds one of these strings, it points at a model that no longer exists on OpenAI's API.
Search your repositories and automation for sora-2 and replace the string with a model you can actually call. For the wider timeline across DALL-E, Sora and GPT Image, see the OpenAI media shutdown calendar.
What Sume offers instead
Sume does not serve a Sora model. Its video API is asynchronous: you list models with GET /v1/videos/models, and each row says which resolutions, durations and frame types it supports. The docs' example row, for seedance-2, shows supported_frame_images of first and last frame and supported_input_references of image, video and audio URL.
Read the live catalog before you copy an id from an example, because the list changes. Then submit with POST /v1/videos, store the job id, and poll the standard job endpoints described in Jobs and results. The Video generation docs cover frame_images and input_references.
A swap is not a drop-in
A different model means different durations, aspect ratios and prompt behaviour, so treat the replacement as a new integration and rerun your own test prompts. Seedance vs Sora 2 compares specs, and the Sora migration post covers the move to Sume's catalog. If you still need clips made with Sora, how to download Sora videos covers what is left.
Sources
Related posts
More in Models
- Sume Auto video has no 1:1 or 21:9: which model to pin instead
Auto in Sume's Videos panel offers 720p or 1080p, 16:9 or 9:16 and 3 to 10 seconds. For square, ultrawide or longer clips, pin Wan 3.0, H3 or Kling 3.0.
- Reference limits per Sume video model: images, clips, audio
How many reference images, videos and audio clips each Sume video model takes on /v1/videos: Seedance 12 total, Wan 10/5/5, H3 9/3/3, Omni 10 and 3.
- Swap the product in a UGC clip with Gemini Omni Flash 1.1 video edit
Keep a UGC clip that works and change only the product: Sume's video router edit mode for gemini-omni-flash-1.1 takes a video_url and a one-line prompt.
- SynthID audio and MP3: does the mark survive Sume's 128 kbps detach?
Google says SynthID audio marks resist noise, MP3 compression and speed changes. Sume audio detach can output 128 kbps MP3; only some sources carry a mark.
Written by Sume