Sora 2 is here

📰 OpenAI News

OpenAI releases Sora 2, a video generation model with improved physical accuracy, realism, and controllability

advanced Published 30 Sept 2025
Action Steps
  1. Explore the capabilities of Sora 2 through the new Sora app
  2. Experiment with prompts to test the model's controllability and physical accuracy
  3. Investigate how Sora 2 can be used for realistic video and audio generation
  4. Analyze the model's ability to obey the laws of physics and model failure
  5. Consider the potential applications of Sora 2 in fields such as animation, film, and video game development
Who Needs to Know This

The Sora team and researchers in AI video generation can benefit from this release, as it provides a more advanced tool for training AI models that understand the physical world

Key Insight

💡 Sora 2's ability to model failure and obey the laws of physics makes it a significant leap forward in video generation capabilities

Share This
🚀 Sora 2 is here! OpenAI's latest video generation model is more physically accurate, realistic, and controllable than prior systems 🤖

Key Takeaways

OpenAI releases Sora 2, a video generation model with improved physical accuracy, realism, and controllability

Full Article

# Sora 2 is here | OpenAI

[Skip to main content](https://openai.com/index/sora-2#main)

[](https://openai.com/)

* [Research](https://openai.com/research/index/)
* Products
* [Business](https://openai.com/business/)
* [Developers](https://openai.com/api/)
* [Company](https://openai.com/about/)
* [Foundation(opens in a new window)](https://openaifoundation.org/)

[Try ChatGPT(opens in a new window)](https://chatgpt.com/)

* Research
* Products
* Business
* Developers
* Company
* [Foundation(opens in a new window)](https://openaifoundation.org/)

[Try ChatGPT(opens in a new window)](https://chatgpt.com/)

OpenAI

Table of contents

* [Deployment of Sora 2](https://openai.com/index/sora-2#deployment-of-sora-2)
* [Launching responsibly](https://openai.com/index/sora-2#launching-responsibly)
* [Sora 2 availability and what’s next](https://openai.com/index/sora-2#sora-2-availability-and-whats-next)

September 30, 2025

[Research](https://openai.com/news/research/)[Release](https://openai.com/research/index/release/)[Product](https://openai.com/news/product-releases/)

# Sora 2 is here

Our latest video generation model is more physically accurate, realistic, and more controllable than prior systems. It also features synchronized dialogue and sound effects. Create with it in the new Sora app.

Loading…

Share

Today we’re releasing Sora 2, our flagship video and audio generation model.

The [original Sora model⁠](https://openai.com/index/video-generation-models-as-world-simulators/) from February 2024 was in many ways the GPT‑1 moment for video—the first time video generation started to seem like it was working, and simple behaviors like object permanence emerged from scaling up pre-training compute. Since then, the Sora team has been focused on training models with more advanced world simulation capabilities. We believe such systems will be critical for training AI models that deeply understand the physical world. A major milestone for this is mastering pre-training and post-training on large-scale video data, which are in their infancy compared to language.

Prompt: figure skater performs a triple axle with a cat on her head

With Sora 2, we are jumping straight to what we think may be the GPT‑3.5 moment for video. Sora 2 can do things that are exceptionally difficult—and in some instances outright impossible—for prior video generation models: Olympic gymnastics routines, backflips on a paddleboard that accurately model the dynamics of buoyancy and rigidity, and triple axels while a cat holds on for dear life.

Prompt: a guy does a backflip

Prior video models are overoptimistic—they will morph objects and deform reality to successfully execute upon a text prompt. For example, if a basketball player misses a shot, the ball may spontaneously teleport to the hoop. In Sora 2, if a basketball player misses a shot, it will rebound off the backboard. Interestingly, “mistakes” the model makes frequently appear to be mistakes of the internal agent that Sora 2 is implicitly modeling; though still imperfect, it is better about obeying the laws of physics compared to prior systems. This is an extremely important capability for any useful world simulator—you must be able to model failure, not just success.

The model is also a big leap forward in controllability, able to follow intricate instructions spanning multiple shots while accurately persisting world state. It excels at realistic, cinematic, and anime styles.

Prompt: Vikings Go To War — North Sea Launch (10.0s, Winter cool daylight / early medieval)...

As a general purpose video-audio generation system, it is capable of creating sophisticated background soundscapes, speech, and sound effects with a high degree of realism.

Prompt: Two mountain explorers in bright technical shells, ice crusted faces, eyes narrowed with urgency shout in the snow, one at a time

You can also directly inject elements of the real world into Sora 2. For example, by observing a
Read full article → ← Back to Reads

Related Videos

5 Levels of AI Agents - From Simple LLM Calls to Multi-Agent Systems
5 Levels of AI Agents - From Simple LLM Calls to Multi-Agent Systems
Dave Ebbelaar (LLM Eng)
AI That Turns Any Concept Into a Tutorial Video (Gemini Omni Flash & Nano Banana II Lite)
AI That Turns Any Concept Into a Tutorial Video (Gemini Omni Flash & Nano Banana II Lite)
Prompt Engineer
GPT-5.6 Sol is HERE — and it Changes Everything (Terra & Luna too!)
GPT-5.6 Sol is HERE — and it Changes Everything (Terra & Luna too!)
Prompt Engineer
GLM_5-2
GLM_5-2
Hyperstack
LongCat 2.0: N-Grams Beat More Experts
LongCat 2.0: N-Grams Beat More Experts
Prompt Engineering
Sonnet 5, more expensive than opus?
Sonnet 5, more expensive than opus?
Prompt Engineering