Comparison
Sora Alternatives: Video Models You Can Use After the Shutdown
Sora's web app was discontinued on April 26, 2026 and its API was scheduled to end on September 24, 2026. Here is what OpenAI says, and nine text-to-video and image-to-video models you can use now, with clip lengths, resolutions and credit prices.
Published November 6, 2026
Sora is no longer available to use. OpenAI's help center says the Sora web and app experiences were discontinued on April 26, 2026, and OpenAI's developer deprecations page lists the Videos API and the sora-2 and sora-2-pro models for removal on September 24, 2026, a date that has passed. For the same job, text-to-video and image-to-video, the current choices include Veo 3.1, Kling 3.0, Seedance 2.0 and 2.5, Wan 3.0 and Grok Imagine Video 1.5. All of them run on reRAW AI, and the table below lists what each one offers.
What happened to Sora
OpenAI's deprecations page records a notice to developers dated March 24, 2026, covering the Videos API and the Sora 2 model aliases and snapshots, with removal on September 24, 2026. Its help center article "What to know about the Sora discontinuation" says the web and app experiences were discontinued on April 26, 2026 and recommends exporting your Sora content as soon as possible. On September 6, 2026 an OpenAI support reply in the developer community said OpenAI had no change to that plan to share. We have not called the Sora API ourselves, so the September date is OpenAI's schedule and not something we tested.
What to look for in a replacement
Four settings differ most between video models and decide what you can make: what the clip starts from (text, a photo, or reference images and clips), how long it runs, the resolution, and whether sound comes with the picture. Cost follows length and resolution, so it helps to know the price of a short test clip before you plan a longer one.
Video models on reRAW AI that cover the same ground
reRAW AI is an all-in-one AI video and image studio. Trending effects turn a single photo into a video or picture in one tap: change outfits, turn a portrait into a cartoon, make a photo dance. For brands, it makes video ads, UGC-style ads, posters and product photos. For full control, the pro studio has top models like Veo, Seedance, Kling and Nano Banana, with reference-to-video, motion control, extend and upscale, plus face swap and background, object and watermark removal. Videos run up to 4K and 30 seconds, half are ready in under a minute, and downloads carry no visible watermark. It works in the browser in eight languages and is free to start.
The table is our catalog on Oct 19, 2026. The credit column is the price of the smallest standard clip, 5 seconds at 480p unless noted, including promotions running that day. Veo 3.1 and Kling 3.0 are priced per clip and by tier, so their model pages show the current number.
| Model | Starts from | Clip lengths | Resolutions | Credits, smallest clip |
|---|---|---|---|---|
| Veo 3.1 (also Fast and Lite) | Text or photo | 4, 6 or 8 s | 720p, 1080p, 4K | See model page |
| Kling 3.0 (Std and Pro) | Text or photo | 3 to 15 s | 720p or 1080p | See model page |
| Seedance 2.5 | Text, photo or references | 5 to 30 s, in steps of 5 | 480p, 720p, 1080p | 110 |
| Seedance 2.0 Mini and Fast | Text, photo or references | 5, 10 or 15 s | 480p, 720p | 12 (Mini), 30 (Fast) |
| Wan 3.0 | Text, photo or references | 5, 10 or 15 s | 480p, 720p, 1080p | 32 |
| Grok Imagine Video 1.5 | Text or photo | 5, 10 or 15 s | 480p, 720p | 15 |
| MiniMax H3 | Photo only | 5, 10 or 15 s | 480p, 768p, and 1440p or 4K upscaled | 14 |
| Seedance 1.5 Pro | Photo | 6 or 12 s | 480p, 720p, 1080p | 8 (6 s clip) |
| Wan 2.6 Flash | Photo only | 5, 10 or 15 s | 720p, 1080p | 12 (5 s at 720p) |
Veo 3.1 is Google DeepMind's model, and dialogue, effects and ambient sound are generated together with the picture. It comes in three tiers, and the larger ones reach 4K. Kling 3.0 is the one to look at when you want to pick the exact length, because it moves in 1-second steps from 3 to 15 seconds.
Seedance 2.5 has the longest single clip on the list, up to 30 seconds, with audio produced together with the picture. Seedance 2.0 takes the widest mix of inputs: text, photos, video clips and audio can all steer one prompt, and it makes clips up to 15 seconds. Wan 3.0 covers text, photo and reference starts at 480p to 1080p, with a sound switch.
Seedance 2.0 Mini and Grok Imagine Video 1.5 are the lowest-priced text-capable models in the table with a listed price, at 12 and 15 credits for 5 seconds. MiniMax H3, Seedance 1.5 Pro and Wan 2.6 Flash start only from a photo, so they suit the case where you already have the first frame. For moving a performance from one video onto a photo, Kling 2.6 Motion Control takes a driving video of 3 to 30 seconds.
Moving your prompts over
We have not tested Sora prompts on these models, so we cannot say how closely a saved prompt carries across. What we do have is a July 2026 series of dance tests on Grok Imagine Video. A prompt split into timed segments with one main action each was performed more reliably than a flowing paragraph. When the start image already shows the person and the place, describe only what moves. The longer write-up is in image-to-video prompts that work.
Choosing
Start with Veo 3.1 for built-in dialogue and ambience, Seedance 2.5 for a long single clip, Kling 3.0 for fine control of length and Seedance 2.0 for mixing text, photos, clips and audio in one prompt. For a cheap first test of an idea, use Grok Imagine Video 1.5 or Seedance 2.0 Mini. The same short prompt on two of them tells you more than reading specs.