Sora 2
Sora 2 is a video generation model from OpenAI that creates videos from text prompts, images, and character references.
OpenAIText and image to video generation by OpenAI
Sora 2 is a video generation model developed by OpenAI, released in September 2025. It accepts text prompts, input images, and character reference videos to produce video output, and is priced at $0.10 per second of generated video. The model supports a context window of 5,000 tokens and allows users to configure output duration and resolution at generation time.
Sora 2 is designed for use cases that require generating video content from descriptive text or visual references, including character-consistent video creation using a named character reference workflow. By accepting both image URLs and character reference video URLs as inputs, it enables workflows where visual consistency across generations is a priority. Developers and creators can integrate it into applications via MindStudio without managing API keys directly.
What Sora 2 supports
Text-to-Video
Generates video clips from natural language text prompts using a 5,000-token context window to interpret detailed scene descriptions.
Image-to-Video
Accepts an input image URL and animates or extends it into a video clip, allowing visual content to serve as a generation starting point.
Character Reference
Takes a reference video URL and a character name to maintain visual consistency of a specific character across generated video outputs.
Configurable Output Size
Supports selectable output resolution and aspect ratio at generation time, giving control over the dimensions of the produced video.
Configurable Duration
Allows users to select the duration of the generated video clip, with pricing applied at $0.10 per second of output.
Ready to build with Sora 2?
Get Started FreeCommon questions about Sora 2
How is Sora 2 priced?
Sora 2 is priced at $0.10 per second of generated video output. The total cost of a generation depends on the duration you select for the output clip.
What inputs does Sora 2 accept?
Sora 2 accepts text prompts, an optional input image URL, an optional character reference video URL, a character name, and configuration selections for duration and output size.
What is the context window for Sora 2?
Sora 2 has a context window of 5,000 tokens, which applies to the text prompt describing the video content to be generated.
When was Sora 2 released?
Sora 2 was released in September 2025 and became available on MindStudio in October 2025.
What is the character reference feature used for?
The character reference feature allows you to supply a video URL of a character along with a character name, enabling the model to maintain visual consistency for that character in the generated output.
Parameters & options
Optional URL of an input image to animate. Must be the exact dimensions as the video output.
(Optional) Upload a 2-4 second video of objects, animals, or animated characters as references to appear within your videos to maintain recognizable mascots and products across campaigns and creative assets.
(Optional) Name for the character (must be mentioned in the prompt)
Explore similar models
OpenAISora 2 Pro
Sora 2 Pro is a flagship video generation model from OpenAI that accepts text, image, and video inputs to produce video output.
OpenAIGPT-6 Astra
OpenAIGPT Image 2.5 Flare
ByteDanceSeedance 2.5 Turbo
Seedance 2.5 Turbo is a video generation model from ByteDance that supports text-to-video, image-to-video, and video editing modes.
Black Forest LabsFLUX 3 Video
FLUX 3 Video is a video generation model from Black Forest Labs that supports audio, keyframes, and video continuation.
X.aiGrok Imagine 1.5
Grok Imagine 1.5 is a video generation model from X.ai that accepts text, image, and video inputs to produce generated video output.
Start building with Sora 2
No API keys required. Create AI-powered workflows with Sora 2 in minutes — free.