Sora Alternative: Where to Make AI Video Now That Sora Is Gone
OpenAI announced on 24 March 2026 that it would discontinue Sora. The Sora app and website closed on 26 April 2026, and the Sora API followed on 24 September 2026. If you made videos there, you need a new place for the same work; these nine Sora alternatives cover social clips, cinematic shots, open models you can keep, and whole films.
October 2026
What happened to Sora, and what to use instead
OpenAI discontinued Sora in two steps: the web and app experiences ended on 26 April 2026 and the API on 24 September 2026, according to OpenAI's help centre. The closest replacements for prompt-to-video with sound are Google's Gemini Omni Flash in Flow, Kling 4.0, Seedance 2.5 on Dreamina and Runway Gen-4.5. For quick social clips in an app, Grok Imagine. For a whole story rather than one clip, filmaris.
Sora alternatives at a glance
| Tool | Best for | How you pay | Watch out for |
|---|---|---|---|
| Gemini Omni Flash (Google Flow) | Prompt-to-video with consistent characters and voices | Google AI subscription; free Shorts remixes in YouTube | Google's own models only |
| Kling 4.0 (Kling AI) | Long single shots | Free Basic plan, subscriptions, credit packs | Free credits are promotional |
| Seedance 2.5 (Dreamina) | Reference control and editing a clip | Trial credits, subscriptions | Trial credits vary by region |
| Runway Gen-4.5 | Control, editing and professional export | One-time free credits, then subscriptions | Standard and Pro credits reset every cycle |
| Grok Imagine Video 1.5 | Quick social clips with speech | Grok apps; pay per use via API | Up to 720p |
| MiniMax H3 (Hailuo AI) | Video with stereo sound from mixed inputs | Bonus credits, subscriptions, credit packs | Commercial use needs a paid plan |
| Luma Ray3.2 (Dream Machine) | HDR-capable footage, many models in one app | Subscriptions with monthly credits | No free plan listed |
| LTX-2.5 | A model nobody can switch off | Open weights, your own GPU or hosted API | Large download; licence terms above a revenue threshold |
| filmaris | Whole films with story, voices and music | Free first film, plans with credits, or your own fal.ai key | Not for one quick clip; AI video needs a paid plan |
1Gemini Omni Flash in Google Flow: the broadest replacement
Google launched Gemini Omni Flash in May 2026 as a model that takes text, images, audio and video as input and outputs video. In Flow, Google's filmmaking tool, it keeps a character's identity and voice across scenes; the 1.1 update in August added scene extension, start and end frames and 4K upscaling. It comes with Google AI Plus, Pro and Ultra subscriptions, and developers pay per use on the Gemini API. There is also a free way in: remixing Shorts with Omni costs nothing in YouTube and the YouTube Create app. Every Omni video carries Google's SynthID watermark.
Best for: Sora users who want prompt-to-video with sound and consistent characters
- Identity and voice kept across scenes
- Free Shorts remixes in YouTube
- Scene extension and start and end frames
- Full Flow access needs a paid Google AI plan
- SynthID watermark on every video
2Kling 4.0: half-minute shots in one generation
Kuaishou opened Kling 4.0 to users on 15 October 2026. It generates up to 30 seconds in a single pass, takes up to 15 reference assets through Omni Reference and follows up to 10 keyframes, which gives you more say over how a shot develops than a single prompt. Kling also offers an MCP server and a command-line tool for AI agents. The Basic plan is free but has no fixed monthly credits; free credits come with promotions, membership credits last one month and purchased packs two years. Watermark-free downloads are a paid-plan feature. Kling 3.0 remains available too.
Best for: long, directed single shots
- Up to 30 seconds per generation
- References and keyframes for control
- Credit packs valid for two years
- Free credits are promotional, not guaranteed
- Watermark removal needs a paid plan
3Seedance 2.5 on Dreamina: edit instead of starting over
ByteDance launched Seedance 2.5 on 31 July 2026, first on its platform Dreamina. The release centres on longer generations, stronger control through reference assets and targeted editing, so a shot that is almost right can be corrected rather than generated again from scratch. Dreamina also creates images and talking avatars. Eligible users may receive trial credits without a credit card, with amounts that vary by account and region; paid plans add monthly credits, watermark-free downloads and higher resolution. Seedance 2.5 is also available on Runway, Luma and Pika.
Best for: refining a shot through references and edits
- Targeted editing inside a clip
- Reference control
- Available on several platforms
- Trial credits vary by account and region
- Watermark-free downloads are paid
4Runway Gen-4.5: a studio with editing and export
Runway's Gen-4.5 has been its flagship text- and image-to-video model since December 2025, with a first-frame input and HDR output formats added since. Aleph 2.0, released in May 2026 with the Edit Studio app, changes existing footage, for clips up to 30 seconds and across several cuts at once, and Pro plans and above export ProRes 4444 or PNG sequences for grading. Runway's plans also include Kling 3.0 and Seedance 2.5. The Free plan is a one-time credit deposit that does not expire; Standard, Pro and Max add monthly credits.
Best for: creators who edit and grade their footage
- Edits existing footage with Aleph 2.0
- Professional export formats
- Kling and Seedance on the same plan
- The free deposit is one-time
- Standard and Pro credits reset each cycle
5Grok Imagine Video 1.5: quick clips inside an app
If you used Sora as an app on your phone, Grok Imagine is the nearest habit: xAI's Grok Imagine Video 1.5 Fast runs in the Grok web, iOS and Android apps, and the full 1.5 model has been generally available through the Imagine API since 17 June 2026. It turns a picture or text into video with sound generated in the same pass, including background music, effects and lip-synced dialogue, and it can extend a clip from its last frame. Clips run up to 15 seconds at 480p or 720p. Through the API, and on fal.ai, you pay per use.
Best for: short social clips with speech, made on the phone
- Lip-synced dialogue in one pass
- Mobile apps for iOS and Android
- Clip extension from the last frame
- Resolution up to 720p
- Clips up to 15 seconds
6MiniMax H3 in Hailuo AI: sound and picture together
MiniMax released H3, also called Hailuo 3.0, on 31 July 2026. It reads text, images, video and audio as one context and generates clips of up to 15 seconds at up to 2K, with native stereo sound covering voice, effects and music, the kind of synchronised audio Sora 2 had introduced. You use it in the Hailuo AI app or through MiniMax's API, and it is also on Pika and fal.ai. New users start with bonus credits; a subscription adds monthly credits that do not roll over, watermark-free downloads and priority generation. Commercial use starts with the Starter plan.
Best for: clips where voice, effects and music matter
- Native stereo audio
- Text, image, video and audio as input
- Bonus credits for new users
- Clips up to 15 seconds
- Commercial use needs a paid plan
7Luma Ray3.2: one app, several model families
Ray3.2 is the newest Ray model on Luma's Dream Machine, listed next to Ray3.14. The Ray3 family introduced HDR output and a draft mode for quick previews, and Ray3.14 added native 1080p with faster generation. Dream Machine also hosts Seedance 2.0 and 2.5, Kling 3.0 Omni and Gemini Omni Flash, so if one model disappears, the next is in the same app. Plans are Plus, Pro and Ultra with monthly credits, plus Team and Enterprise, and commercial use starts with Plus. Luma's pricing page lists no free plan.
Best for: comparing several model families in one place
- HDR output in the Ray3 family
- Seedance, Kling and Gemini Omni on the same plan
- Commercial use from the first plan
- No free plan listed
- Credits only with a subscription
8LTX-2.5: open weights that stay yours
The Sora shutdown showed that a hosted model can disappear along with your workflow. Open weights cannot be switched off once you have them. Lightricks released LTX-2.5 as open weights on 11 August 2026; it generates picture and synchronised audio in one pass and runs on your own GPU, in ComfyUI or through Lightricks' hosted LTX API. The model files are roughly 66 GiB, and the LTX-2.x Community License is free for smaller organisations but asks companies above a revenue threshold to negotiate a licence. For a fully open alternative without that threshold, Wan 2.2 is available under Apache 2.0.
Best for: a video model with sound that no provider can withdraw
- Weights stay on your own hardware
- Video and sound in one pass
- Free for smaller organisations
- Large download and a strong GPU
- Revenue threshold in the licence
9filmaris: a whole film, not tied to one model
filmaris is the right pick if you used Sora for stories rather than single clips. It starts from one sentence, a screenplay or a book, writes the script, gives every character, location and prop a reference sheet, and plans each shot. Each shot then goes to the model that suits it among more than fifty, including Kling 3.0, Veo 3.1, Seedance 2.5, Hailuo 2.3 and Wan 2.7, so the film does not depend on any single provider. Voices, music and sound effects are added, and you edit on a multitrack timeline and export MP4 up to 4K or FCPXML. For one quick clip or a talking avatar, other tools here fit better. Your first one-minute storyboard film, with camera moves and voiced dialogue, is free.
Best for: complete films with story, recurring characters, voices and music
- More than fifty models, chosen per shot
- Reference sheets keep characters consistent
- MP4 and FCPXML exports you keep
- Not built for single clips or avatars
- Real AI video shots need a paid plan
How we compared
We confirmed the Sora timeline in OpenAI's help centre and in reporting by Engadget and The Decoder, then looked at the providers former Sora users name most often. For each entry we read the launch announcement, pricing page or documentation between 16 and 22 October 2026 and kept a source for every statement. Prices are left out because they change often and vary by region; we compare how you pay instead. We ran no paid test generations for this article. filmaris is our own product and is listed for the job it is built for.
Bring your stories to a new home
Describe the film you wanted to make in one sentence. filmaris writes it, casts it and plans every shot, and your first one-minute storyboard film is free.
Start your film freeSora shutdown and alternatives: common questions
Is Sora shut down?
Yes. OpenAI announced on 24 March 2026 that it would discontinue Sora. The Sora web and app experiences ended on 26 April 2026 and the Sora API on 24 September 2026. OpenAI's help centre says Sora data is permanently deleted after these dates and after any final export window.
Can I still download my Sora videos?
OpenAI asked users to export their videos and images before the cut-off dates and said it would e-mail users if a final export window is offered. After that, the data is deleted. Check OpenAI's help article on the Sora discontinuation for the current status.
What is the best free Sora alternative?
Remixing Shorts with Gemini Omni costs nothing in YouTube and the YouTube Create app. Kling's Basic plan, Dreamina's trial credits, Hailuo AI's bonus credits and Runway's one-time deposit let you test without paying. Wan 2.2 and LTX-2.5 run free on your own graphics card. filmaris gives each confirmed e-mail one free one-minute film made from storyboard pictures with voiced dialogue.
Which Sora alternative generates sound with the video?
Several do. Gemini Omni Flash keeps characters' voices across scenes, MiniMax H3 generates native stereo audio, Grok Imagine Video 1.5 adds music, effects and lip-synced dialogue, and LTX-2.5 creates picture and sound in one pass. filmaris adds voices, music and sound effects to a whole film and mixes them on its timeline.
How do I avoid losing my tool again?
Keep your source material and finished exports on your own drive, prefer formats other editors can open, and do not depend on a single model. Open weights such as Wan 2.2 and LTX-2.5 keep running on your machine. filmaris chooses among more than fifty models per shot and exports MP4 and FCPXML, so a film is not tied to one provider.