
OpenAI shuts down Sora: a bombshell in the video AI world

Reading time: 5 min
This is a shock that few observers expected. Just a few months after generating worldwide excitement, Sora, OpenAI's video generator, is bowing out. Even more surprising, this decision comes at a time when the tool was enjoying considerable success among creators, agencies, and marketing teams.
For many users, Sora embodied a promise: democratize video production, reduce costs and accelerate the creation of content at scale. Its sudden disappearance leaves a void, but also many questions about the economic and strategic viability of these technologies.
A technical success… but a fragile model
At its launch, Sora made a strong impression with its ability to generate realistic videos from simple text prompts. The tool was part of a dynamic of industrializing content creation.
However, behind this technical feat lay a more complex reality. Video generation is extremely resource-intensive. Unlike text models or even image generators, video involves much larger volumes of data, longer computation times and a significantly more demanding infrastructure.
This imbalance between rapid adoption and high operating costs probably weighed heavily in the balance.
The reasons for the shutdown: between financial constraints and strategic choices
According to experts in the field, several factors may explain OpenAI's decision to discontinue Sora.
From a financial point of view, the economic model remained difficult to stabilize. Each video generated represents a significant expense., difficult to compensate for even with premium subscriptions.
Strategically, OpenAI seems to be refocusing its efforts on more profitable and more widely disseminated products. Video remains an unstable market that is difficult to standardize.
Finally, regulatory and ethical issues may also have played a role. The generation of realistic videos raises major questions about deepfakes and disinformation.
What alternatives are there for creators and marketing teams?
Sora's disappearance does not mean the end of generative video. The market remains active, but fragmented. Several players continue to invest heavily in this area.
Here is a broader overview of the main alternatives available today:
| Last Name | Monthly cost (indicative) | Description |
|---|---|---|
| Runway ML | ~15 to 35 € | Advanced AI-powered video generation and editing platform, geared towards creators and studios |
| Pika Labs | Free + premium offers | Video generation from text, with a strong creative focus |
| Synthesia | ~€30+ | Creating videos with AI avatars, widely used in training and corporate communication |
| Hey Gen | ~€24+ | Video generation with avatars and synthetic voices, for marketing and corporate use |
| Luma AI (Dream Machine) | Freemium | Realistic video generation with an emphasis on visual quality and fluidity |
| kaiber | ~10 to 30 € | A tool geared towards artistic creation and generative video clips |
| Google Veo | Not disclosed / experimental | Google's advanced video model, geared towards realism and long storytelling |
| Stable Video Diffusion | Open source / variable | Model derived from Stability AI, allowing the generation of short video sequences |
| Adobe Firefly Video | Includes Creative Cloud (beta) | Video generation integrated into the Adobe ecosystem, with strong creative control |
| midjourney | ~10 to 60 € | Initially focused on images, it explores animation capabilities and video sequences. |
| Claude (Anthropic) | ~€20+ | Not a native video generator, but it can orchestrate creative workflows including video |
These solutions adopt a variety of positions. Some aim for realistic production, others prioritize formats controlled or artistic.
A still unstable market
Sora's ending highlights an often underestimated reality: The generative video market is still immature.
On one hand, the demand is very real. Brands have a growing need to produce more content, faster.
On the other hand, the technical and economic constraints remain significant. Infrastructure costs remain a major obstacle to widespread adoption.
The question that now arises is simple: Will other players be able to achieve a viable economic model?
It is likely that the market will experience a consolidation phase in the coming months. Not all the players will survive.
What this changes for marketing strategies
For marketing teams, the disappearance of Sora is an important signal.
She points out that it is risky to depend on a single tool or platform. Diversification is becoming a strategic issue.
At the same time, she confirms that generative video is not a fad. Uses continue to develop rapidly.
Conclusion
The end of Sora marks a turning point in the recent history of generative AI. It reveals both the power and the fragility of these technologies. While the technology remains impressive, its industrialization still poses many challenges. The market is now entering a phase of maturity.
More than an end, the end of Sora could well be the beginning of natural selection in the ecosystem of generative video AI.
Some references
- « Sora: OpenAI's video generator that's revolutionizing content creation "- martech Cloud – March 2025
- « What's next for generative video » – MIT Technology Review – 2025
- « Innovation Insight: AI Video Generation Reshapes Content… – Gartner – Nov 2025
- « Generative AI: the risks and the unknowns – OECD.AI » – OECD – 2024















