Sora shutdown and access
Discontinuation of the Sora app, web product, and API; shutdown dates, preserving creations, and creators losing access.
52%
Best tweets about Sora
Explore the best tweets about OpenAI Sora, including video model releases, prompts, storyboards, visual quality, access, safety, and creator experiments.
OpenAI Sora video generation, product updates, prompts, storyboards, visual results, access, safety, comparisons, and creator workflows.
Original Xholic analysis
Sora’s shutdown dominates this conversation: its team announced the app’s end and promised guidance on preserving work, while posts debated compute priorities and the loss of creator tools (2036546752535470382, 2039863053588845002, 2031561991593275896). Rival-model tests and creator accounts show continued interest in AI video, but not agreement on which product best serves creators (2044400358718386493, 2039027015127531984).
30% of posts
All-time engagement
68% of posts
Published in 90 days
Conversation map
Discontinuation of the Sora app, web product, and API; shutdown dates, preserving creations, and creators losing access.
52%
Sora’s operating costs and user retention, and OpenAI’s decision to redirect compute and staff toward frontier models, coding, enterprise, and robotics research.
36%
References, reusable characters and styles, Cameo, remixing, prompt assistance, editing needs, short films, and rapid ad production.
26%
Same-prompt comparisons with Seedance, Kling, PixVerse, Veo, and Grok Imagine; rival pricing, adoption, and creators moving to alternatives.
26%
Prompt adherence, textures, camera movement, character consistency, physical realism, and scientific reasoning in generated video.
24%
Debate over AI-video engagement, attention scarcity, repetitive ‘slop,’ and the role of humor, taste, and human craft.
14%
The canceled character-licensing plans and proposed investment, including Disney’s reported short notice of Sora’s closure.
14%
Generation guardrails, licensed content and copyright concerns, and deceptive AI-generated footage or deepfakes.
10%
Tone and stance
Performance benchmark
Posts with media make up 64% of this collection. Their median all-time score is 5.40, compared with 9.05 for text-only posts.
Format mix
Consensus and debate
Shared view
Sora’s team announced the app’s end and said app and API timelines and preservation details would follow. Sora promoted reusable references, while creators singled out remixing and Cameo as features they valued and may miss.
Shared view
Altman described concentrating compute and researchers on next-generation models and agents; Brockman likewise discussed difficult compute allocation. Posts citing reporting added claims about Sora’s costs, which remain attributed claims rather than independently established here.
Open debate
Some creators praised Sora’s humor, remixing and short-film possibilities; others argued that repetitive AI clips cannot substitute for taste, craft or scarce audience attention.
Open debate
Same-prompt reviewers praised Sora 2 Pro’s textures but flagged motion, realism or prompt-adherence issues. A former user nevertheless preferred its quality to Grok’s, while another reported friends preferred Grok Imagine’s API.
Open debate
One post questioned whether the shutdown reflected costs, copyright or product execution despite popular AI clips. A creator reported frustrating generation blocks; a fact-check traced deceptive footage to channels whose earlier uploads carried Sora watermarks.
What performs
The team’s goodbye post scores 3445.47, or 540.89 times the dataset median of 6.37. The supplied analytics classify shutdown and access as the largest theme: 26 of 50 tweets, or 52%.
The rival-model theme has a supplied median all-time score of 10.692, versus 5.86 for shutdown and access. A Seedance-versus-Sora same-prompt post is among the supplied score outliers; this does not establish which model performs best.
Statistical standouts
Creator landscape
The five most represented creators account for 18% of the selected posts.
1. Andrew Curran
@AndrewCurran_
2 posts
2. Roope Rainisto
@rainisto
2 posts
3. Sora
@soraofficialapp
2 posts
4. VraserX e/acc
@VraserX
2 posts
5. Jumanne Mtambalike
@Afruturist
1 post
6. AI Frontliner
@AIFrontliner
1 post
Sora promoted saved characters, styles, props and camera moves; creators also valued feed-based remixing and prompt assistance. These accounts frame the product as a creative workflow, not just a text-to-video model.
Creators described rapid ad variations and short films, but an early user identified cuts, subtitles and hooks as missing production tools. Another creator hoped reference images in Seedance might replace part of a Sora editing workflow.
Since the previous snapshot
Themes, sentiment, stance, and post format are classified per tweet. All counts, shares, medians, creator concentration, freshness, and performance comparisons are then calculated directly from the published snapshot.
Xholic's all-time score compares engagement while accounting for reach, post age, and creator consistency. It is used for relative comparisons within this collection.
This report analyzes the exact 50-post snapshot shown below. AI identifies editorial categories and drafts explanations; all statistics are calculated from the snapshot, and every narrative claim is checked against cited posts before publication.
Best Sora tweets
Ranked 01–50
@soraofficialapp ·
We’re saying goodbye to the Sora app. To everyone who created with Sora, shared it, and built community around it: thank you. What you made with Sora mattered, and we know this news is disappointing. We’ll share more soon, including timelines for the app and API and details on preserving your work. – The Sora Team
@AndrewCurran_ ·
Some highlights from the WSJ Death of Sora article: - Disney execs learned about Sora being shut down less than an hour before it was announced, this verifies the reporting from Reuters - OpenAI needed to free up compute for Spud, specifically for enterprise tool use - The training run for Sora 3 was about to begin, and the estimated cost was very high - OpenAI had begun piloting an enterprise version of Sora for release in the Spring - at the time of its cancellation Sora had less than 500k users, and was losing a million dollars a day - Disney is currently in active discussions with more than a dozen different companies for their new AI video model and AI tools
@tonjkb ·
Sam Altman explains why OpenAI decided to shut down Sora "So apparently you're not doing Sora anymore. You guys have made the decision to shut it down. So take me to the room, Sam. It like three months ago openai signed a deal with Disney This is like a landmark deal in so many ways Disney going to license 200 characters Pixar Marvel Star Wars Everybody has an opinion on this Bob Iger, who led the deal, he said this, not me. We're bringing together Disney's iconic stories and characters with OpenAI's groundbreaking technology. They're going to invest a billion dollars. Fast forward three months later, OpenAI shuts down Sora." "What's the note under the note here?" "We have a few times in our history realized something really important is working or about to work so well that we have to stop a bunch of other projects in fact this was the original thing happening with gpt3 we had a whole portfolio of bets at the time a lot of them were working well we shut down many projects that were working well like robotics which we mentioned so that we could concentrate our compute our researchers our effort into this thing that we said okay there's a very important thing happening um i did not expect three or six months ago to be at this point we're at now where something very big and important is about to happen again with this next generation of models and the agents they they can power but i love sora i love generated videos and i love our partnership with disney" @sama @OpenAI @soraofficialapp
@Pirat_Nation ·
New details from the video generator Sora OpenAI shutdown - It was losing roughly $1 million per day on compute costs. - User numbers had dropped below 500,000 from a peak of over 1 million. - Disney executives received less than an hour's notice despite ongoing talks for a major integration.


@Framer_X ·
Seedance 2.0 vs Sora 2 ⚔️ I tested the same prompt with Ghibli aesthetics. The results speak for themselves. Quick breakdown below 👇
@tszzl ·
sora was a peak moral panic. all of these breathless takes about making videos that are going to addict humanity and waste everyone’s time. meanwhile we made some funny videos that were less funny as time went on and ai slop is just one category among many on instagram reels
@AndrewCurran_ ·
Greg Brockman on why OpenAI felt they had to drop Sora: 'There's been this debate of how far will the text models go? How far can text intelligence go? Can you have a real conception of how the world operates? And I think that we have definitively answered that question. It is going to go to AGI. Like, we see line of sight. And at this point we have line of sight to these much better models that are coming this year. And the amount of pain within OpenAl that we've had to decide how to allocate compute, that goes up not down over time. So I think that maybe the core of it is ... in this moment the kinds of applications that we've always dreamed of are starting to come into reach.'
@soraofficialapp ·
references on sora. save characters, styles, props, costumes, camera moves, or prompt cheat sheets then reuse them across videos to keep your work consistent. find the new references tab in the video composer. available now on web for sora users.
@signulll ·
the entire sora chase kinda cost openai twice. on the supply side it bled compute & possibly researcher attention from the frontier. & on perception side, it handed anthropic the moral high ground for free by shipping a consumer deepfake product with a feed at the exact moment the labor displacement discourse peaked which effectively signaled less seriousness in many ways esp for business products/customers. you can see the latter in how dario mentions "the slop we've seen with ai video models” explicitly in his latest interview. (i just finished watching the long form dario bloomberg piece).
@himanshustwts ·
OpenAI has to be one of most volatile org right now. Just now, Head of Sora and VP of Science announced their departures. I believe OpenAI is narrowing down what matters the most right now - Coding, Enterprise and Frontier Research. I think this might follow more dec in team size.


@veggie_eric ·
After the Sora API was deprecated last week, this was the forcing function for some of my startup friends to finally try the Grok Imagine API. They told me they liked the video quality even more, and were surprised that it was cheaper and faster🔥🎬
@Similarweb ·
Following the announcement of Sora’s shutdown, here’s how standalone AI video generation traffic evolved over the past 12 months. Sora traffic includes only its official domains (https://t.co/vMIIVocu4Y and https://t.co/8p96loFZ4D). Grok Imagine reflects activity on https://t.co/gaxv74n1YA.

@Skyebrows ·
I'm sure Sora 2 will be replicated/surpassed. But until then I'm going to miss making interdimensional cable trailers. Maybe Seedance 2 can do something similar with enough reference images without having to open the editor.
@frankyecom ·
I love making ads with my 5.1 update This ad would’ve cost like $2-3k to make 1 year ago… Info, product, service, lead gen. You name it this is the framework to use. I pioneered a lot of the UGC realism stack marketers use today with Sora and I think this is even better…
@pleometric ·
I've been getting dms from young IG and TT users asking me how to make AI videos. "What app do you use?" I've talked to few of them and they either don't own a laptop, or they rarely use it. According to their profiles most of them are in the US, so it's not strictly an access issue. This is a very small sample size, like 12 conversations or so but it does add to my growing suspicion that younger cohorts are using short form as a stand in for apps. They "prompt" accounts that post AI content to tune the videos closer to what they want in lieu of having model access. They at best use GPT image gen (and previously SORA), the more savvy ones game free tiers from companies like heygen and etc to make their own variations, which they often find lacking.
@poezhao0605 ·
OpenAI announced today it is shutting down Sora. On the same day, Kuaishou’s Q4 earnings revealed Kling AI generated RMB 340 million in revenue last quarter. ARR surpassed $300 million by January. 60 million global users. 600 million videos created. The divergence is striking. OpenAI treated video generation as a showcase. Kuaishou treated it as a product line.
@cinamarina ·
This is AI and I still can’t believe it 🤯 I just created a UGC-style video in less than 10 minutes. As a solo founder, this kind of speed honestly feels like a cheat code. I used Arcads AI and Sora 2 to create it. I’m starting marketing for my app soon, and TikTok will be one of my main channels, so I started studying competitors. I found a TikTok video from a competitor with 40k+ views, took the hook, and adapted it for my own app. Then I picked an actor in Arcads AI that matched the kind of woman I wanted for the video, generated it in minutes, and the quality is just on another level. Now I know I can create many more variations in just a few minutes. We're entering a completely different era of marketing.
@AIFrontliner ·
Same prompt, four AI video models tested on a carousel scene. Here is how they compare in scene richness, physics, and camera motion. PixVerse V6 delivers smooth motion, rich scenes, matched audio, and realistic textures. However, its camera remains fixed and does not follow the carousel's rotation. The output also drifts slightly from the prompt, trading strict realism for bold imagination. Great creativity, less precision. Sora 2 Pro offers excellent texture quality and cinematic feel. But it also deviates from the prompt, lacks realism, and suffers from mild camera instability. The shaky camera slightly hurts immersion. Seedance 2.0 performs well in the first half: highly realistic, smooth motion, and solid scene understanding. However, the second half degrades, and the overall consistency is uneven. Still, its best moments are impressive. Seedance 1.0 Lite struggles the most. The output is abstract and barely recognisable as a carousel. It fails to understand the basic prompt, resulting in poor scene richness, broken physics, and erratic camera movement. Each model has trade-offs. PixVerse V6 wins in creativity and audio, Sora 2 Pro leads in texture, Seedance 2.0 shows realistic potential in short bursts, and Seedance 1.0 Lite needs more training.
@XFreeze ·
Grok-Imagine just literally overtook the entire video leaderboard on DesignArena Clean sweep...4 out of 4: 🏆 #1 in Video Arena 🏆 #1 in Video-to-Video 🏆 #1 in Image-to-Video 🏆 #1 in Multi-Image-to-Video Outranking Veo 3.1, Sora, Kling - all of them Just a few months ago, xAI wasn't even in the video generation conversation. Now owns the entire space The speed of progress is insanely fast at xAI

@berber_jin1 ·
SCOOP - OpenAI is set to discontinue its Sora video app plus Sora API, the latest step in its efforts to focus the company's resources around enterprise and coding more to come https://t.co/qnOwatHx1C
@venturetwins ·
One of the really cool things about Sora was the remixing culture. It made it super easy for new creators to get started - scroll a feed, see something you like, and riff on it (vs. writing your own long prompt). Hope to see that brought to other AI video products!
@TheAIColony ·
Same prompt, four AI video models tested on a damaged android crawling through a burning spacecraft corridor. Here is how they compare in scene richness, physics, and camera motion. PixVerse V6 delivers smooth motion, rich scenes, matched audio, and realistic textures. The android body is highly realistic with coordinated movements. The model also handles close-up and wide shot transitions naturally. See it here: https://t.co/tohJuo18Nb… Sora 2 Pro offers excellent texture quality and the android body is well constructed. However, the motion is not fluid, making the crawl feel jerky and less convincing. Seedance 2.0 has strong texture quality as well, but the android body looks too human, lacking the mechanical endoskeleton feel. The imagination is somewhat limited. Hailuo-02 also delivers good texture quality, but the android is shaped more like a monster than a damaged biological android, losing the intended sci-fi aesthetic. Each model has strengths and trade-offs. - PixVerse V6 leads in motion smoothness and audio-visual sync. - Sora 2 Pro excels in texture but struggles with motion fluidity. - Seedance 2.0 has solid visuals but lacks mechanical imagination. - Hailuo-02 needs better prompt adherence for character design.
@julesterpak ·
Something’s off. Many people are acting like “the people are fighting back” but the avg person has been embracing generative AI more than ever lately? Like y’all… that AI fruit Love Island series on TikTok has been the most attended to media on the internet over the past week I understand Sora the social media app being shut down bc it had poor product evolution/execution, but I don't understand Sora within ChatGPT going. So much gen Al video comes across people's feeds and (for better or worse) is enjoyed. So is it just the cost? Is it all the copyright headaches? Is it that other platforms are used far more — what company is actually leading when it comes to what people use to generate Al video for social media rn?
@VraserX ·
Looks like Gemini / Veo 4 is getting its own version of Sora 2’s Cameo feature. Honestly, that is genuinely cool. I dearly miss Cameo. Dropping yourself into AI video was one of the most fun and human features Sora had.
@alexabelonix ·
OpenAI is quietly shutting down Sora OpenAI is winding down Sora, the video model that flooded feeds with AI-generated clips just months ago. No clear explanation yet. Just a promise to share details later on how the app and API will be shut down. p.s. that's actually sad because i generated a lot of videos there and the quality is better than Grock!

@rronak_ ·
AI are replacing everything, what if they replace Operating Systems altogether? New paper from Meta, trains model to predict next frames of a computer, and elicits emergent capabilities that emulate an entire OS. The next step is to scale up this training, do post-training, and we could even get fully turing-complete computers. They also find that Sora 2 out of the box is pretty capable of this too. Crazy!! Authors: @MingchenZhuge et al

@ZeffMax ·
It's notable that OpenAI will continue its world/video model research under its robotics org, even tho its nixing the Sora product. It sucks for the Sora team, but it does seem like a positive signal for OpenAI, which has spread itself thin across many projects over the years.
@haoailab ·
Seedance-2 and Kling-3 signal that AI video generation is entering a “photorealistic” era, but realism does not guarantee reasoning and scientific correctness. In the classic breaking dry spaghetti experiment, real fractures arise from elastic energy and stress waves, often producing three pieces. Yet models frequently generate physically incorrect snaps. We introduce VideoScience-Bench to evaluate scientific reasoning in video generation. Most models look convincing but lack true scientific understanding, with only weak signals from Kling-3, Sora-2, and Veo-3. Learn details from our blog 👇🧐🧪 https://t.co/Vm1ncKnm58
@VraserX ·
I actually miss Sora 2. Not just the video quality. Whatever underlying model it used to rewrite and improve prompts had a genuinely great sense of humor. It understood timing, absurdity, deadpan comedy, visual irony. The stuff it came up with often felt funnier than what I originally asked for. That’s rare in AI. Most models can “write jokes,” but they don’t really have taste. Sora 2 somehow did. I genuinely wonder if OpenAI could bring that comedic taste into ChatGPT’s text and voice models. @sama

@flowith ·
ran the same claymation love triangle prompt with sora 2 and kling 3.0 on flowith canvas, then compared with seedance 2.0. honestly...seedance 2.0 is wild...looking foward to bringing it to flowith once available. video and prompt below👇
@rainisto ·
#1 Short movie. Sora 2. -- Creating this, I was thinking that in a few years I would have to work hard for it to come out like this - "you want her identity to be not consistent? You want the environment to change shot to shot?" The sound of the instrument shapes the song you want to play with it. The idea of the story, the mood, came into mind because of the way it is, not despite of it.
@wallstengine ·
OpenAI is planning to discontinue the app for its Sora video platform as the company shifts more of its focus toward business and coding products ahead of a potential IPO later this year. - WSJ

@hq4ai ·
Sora, 25 seconds. Even if everyone makes videos, everyone will still lack traffic. The attention market is finite — more videos don’t mean more viewers. Video content was already overloaded before AI. If you just copy what others do, it means little.

@glenngabe ·
"Sora was a money pit that nobody was using, and keeping it alive was costing OpenAI the AI race" -> Why OpenAI really shut down Sora "So what happened? After a splashy launch, Sora’s worldwide user count peaked at around a million and then collapsed to fewer than 500,000. Meanwhile, the app was burning through roughly $1 million every day — not because people loved it but because video generation is so costly to run. Every user who dropped themselves into a fantastical scene was drawing down a finite supply of AI chips." https://t.co/hpr4HRNr4C
@natlungfy ·
New: Rival AI video generation apps such as Kling AI, RunwayML and Vidu have all seen user gains in the week since OpenAI said it would shutter its Sora AI video generator, per Sensor Tower data https://t.co/ASP5PsTfHp
@mexitlan ·
I am not sure it is appreciated how much of the @OpenAI Sora shuttering is actually a reallocation of video-generating resources away from entertainment and toward physical AI. Even though they literally said: "the Sora research team continues to focus on world simulation research to advance robotics." 🧵
@ZeevFarbman ·
Watching the news about Sora. Hard not to think about how fast this space moves. Performance alone was never going to be enough. The models that matter long term are the ones that are fast, open, and fit inside a real production pipeline.
@rainisto ·
THE RUNNER A woman is out running in the rain. Short movie, made with OpenAI Sora 2 text-to-video. 🧵
@exec_sum ·
BREAKING: After announcing the closing of AI video generation app Sora, Disney has cancelled their deal with OpenAI They planned to invest $1 billion into the company and let Sora have access to AI generate videos of characters from Disney, Marvel, Pixar and Star Wars

@chatcutapp ·
When Sora dropped, the fantasy was: “You’ll never need to shoot anything again.” Turns out people do get tired of slop. Fast. The internet does not need infinite cats with guns and cinematic bunnies. AI or no AI, some things never change: taste matters. intention matters. craft matters. Slop sinks. It always has. The single most important question remains: How do we move AI creative tools beyond slop generators, into tools that actually expand human creativity. AI's got a bad name right now that many filmmakers and creators simply do NOT believe the day will come. But at ChatCut, we live for that day. https://t.co/OXoIrvpUvB
@shawnchauhan1 ·
A billion-dollar deal disappears overnight when a product gets killed. OpenAI shut down Sora to redirect compute toward its next model, codenamed Spud. Disney walked away from a planned $1B investment tied to a Sora licensing deal. The deal was signed in December. It never closed. This is not a product sunset. It is a signal about how disposable consumer AI products have become. When compute is the scarce resource, nothing else is safe. Not partnerships. Not flagship launches. Not brand promises. The AI arms race is now being fought at the infrastructure layer, not the product layer.
@Afruturist ·
OpenAI shuts down Sora as Disney drops $1B investment plan... OpenAI will discontinue its Sora video app. Following the decision, Disney ended its partnership and scrapped plans for a $1B investment. The deal included using Sora to generate videos with licensed characters from Disney, Marvel, Pixar, and Star Wars. Those plans are now canceled. Sora had launched in 2025 as a standalone text-to-video app, but faced industry pushback around copyright and content use.

@GFCNofficial ·
The viral video of a tiger escaping an enclosure is AI-generated Visual analysis reveals multiple AI artifacts: • An unconnected water hose on the floor. • A sliding cage door shifting off its rails. • A background door opening on standard hinges. • Distorted timestamp digits. • Unnatural physical movements during the tiger’s jump. Reverse image searches link this footage to YouTube channels specializing in AI-generated disaster content. Earlier uploads on these accounts feature OpenAI’s Sora watermark, and the creators explicitly state in their bios that the videos are made with artificial intelligence. #FactCheck #Deepfake #AI #Misinformation #Sora

@unable0_ ·
well, sora getting shut down was kind of expected. inference-heavy consumer products with weak retention were always on borrowed time.

@Ldnbox ·
WSJ today published a very interesting piece about OpenAI and Sora. -Disney executives learned about the shutdown less than an hour before it was publicly announced, leaving them blindsided. -A key driver of the shutdown was compute scarcity — Sora consumed enormous amounts of AI chips relative to the revenue it generated. -OpenAI needed to free up computing resources for a new model codenamed "Spud," aimed at powering coding and enterprise products. -Sora was losing approximately $1 million per day at the time of its closure. -Disney's $1 billion investment in OpenAI never closed, and the partnership is now effectively dormant. -Altman framed the shutdown internally as a necessary trade-off, praising staff for their willingness to make difficult decisions for the company's long-term benefit.

@WizzyOnChain ·
JUST IN: Google shook up the generative video landscape by halving the API cost of Veo 3.1 Lite, at a time when OpenAI's Sora offering is leaving the market due to its enormous associated costs. The move points to more aggressive competition among developers, companies, and creators looking to produce AI-powered video at a lower cost. Google reduced the API cost of Veo 3.1 Lite by 50%, reinforcing its commitment to generative video.


@minhsmind ·
OpenAI CFO: “even for things like Sora & video …btw yah we had to make a really touch choice because we didn’t have enough compute. But Video is not over.”
@bearlyai ·
OpenAI has internal dashboard where researchers track compute use. They saw how much compute Sora was using without revenue (losing $1m a day) and moved to shut it down. Focussed scarce compute on latest Spud model.

@shivsakhuja ·
I got access to Sora 2 today and have been playing around a bunch. Initial thoughts: 1. The Cameo feature is amazing and super fun — really good character consistency. Not quite nano banana level, but maybe 80% of the way there. 2. Most videos I’ve seen aren’t actually engaging (yet). TikToks and Reels are much more engaging because they’re longer, better edited, and way more polished. Plus there’s not a ton of content there on Sora yet for the algo to feed you from. 3. Production-grade editing (good cuts, subtitles, hooks, etc) is a key missing ingredient currently. I suspect OpenAI will release really good editing tools in the Sora app itself within a few weeks / months. 4. There are extremely strict guardrails around generations — more than half my generations got blocked for no clear reason and it’s pretty frustrating. There are a lot of people complaining on Sora itself about it too. 5. OpenAI really knows how to productize. This actually feels like it's likely to become a mass consumer app if they feed creators with the tools to make the content more engaging. The creation is so easy and good, it’s addictive to play with. 6. Can't wait for the API
Best Sora tweets
Xholic studies what works in your niche, drafts posts in your voice and schedules them for the hours your audience is online.
$0 today · Cancel anytime
Browse all tweet collectionsKeep exploring