All models
Video generationDeprecated
OpenAISora 2
OpenAI's legacy synced-audio video model for short text- or image-guided clips—still documented, but approaching a permanent API shutdown.
Plain-English overview
What Sora 2 actually is
Sora 2 generates short landscape or portrait video with synchronized audio from text and optional image guidance. Its simple per-second price makes small experiments easy to estimate.
This is now primarily a migration entry, not a recommendation for a new production build. OpenAI has scheduled the Sora API to shut down permanently on September 24, 2026.
Good fit for
- Evaluating an existing Sora integration
- Short synced-audio experiments before shutdown
- Planning a migration to another video provider
Category comparison
The facts that matter for video models
These are provider-published specifications, not Cody benchmark scores. Follow the linked sources for current limits and endpoint-specific exceptions.
- Generation price
- $0.10 / generated secondCurrent list price per generated second or the provider's closest billing unit.
- Clip duration
- 4, 8, or 12 secondsDocumented duration choices or maximum clip length.
- Maximum output
- 720×1280 or 1280×720Highest provider-documented output resolution, with relevant constraints.
- Native audio
- Yes, synced audioWhether the endpoint generates synchronized audio with video.
- Input modes
- Text or image to videoSupported text, image, video, reference, or motion inputs.
- Generation time
- Not publishedOnly shown when a provider publishes a range; not a Cody benchmark.
Pricing & comparisons
Estimate your cost
Set your usage. Your estimate updates as you type.
Example: 8-second, 720p clips with audio. Models that do not support these settings will not receive an estimate.
Sora 2
OpenAI
Estimated total (USD)
For the usage above · USD · API pricing, not a subscription
How this estimate works
Estimates exclude taxes, tools, cache storage/writes, free allowances and custom discounts. Image estimates cover output only, not prompt or reference-image charges. Quality modes differ by model. Unlisted settings are not treated as free.
API and provider access
Where to get Sora 2
Availability
Regions and access stage
Legacy OpenAI Videos API access remains temporarily available, but new integrations should account for the announced shutdown.
Direct access follows OpenAI's supported-country list while the API remains active.
Check live availabilityData and training
The route matters.
OpenAI says API inputs and outputs are not used for training by default. Standard abuse-monitoring and feature-specific storage controls apply until the service shuts down.
This is a concise reading of the cited provider material, not legal advice. A third-party gateway can have different storage, routing, training, and residency terms from the model maker's direct API.
Read the provider policyFrequently asked questions
Sora 2 FAQ
What is Sora 2?
OpenAI's legacy synced-audio video model for short text- or image-guided clips—still documented, but approaching a permanent API shutdown. Sora 2 generates short landscape or portrait video with synchronized audio from text and optional image guidance. Its simple per-second price makes small experiments easy to estimate.
When was Sora 2 released?
Sora 2 was released on December 8, 2025 according to the cited provider materials.
Is Sora 2 still supported?
The Sora API is scheduled to permanently shut down on September 24, 2026. Do not begin a new long-lived integration without a migration plan.
Where can I access Sora 2?
Legacy OpenAI Videos API access remains temporarily available, but new integrations should account for the announced shutdown. The access routes listed in this guide are OpenAI.
How much does Sora 2 cost?
$0.10 / generated second. OpenAI list price for 720×1280 portrait or 1280×720 landscape Sora 2 output.
Where is Sora 2 available?
Legacy OpenAI Videos API access remains temporarily available, but new integrations should account for the announced shutdown. Direct access follows OpenAI's supported-country list while the API remains active.
Is my Sora 2 API data used for training?
OpenAI says API inputs and outputs are not used for training by default. Standard abuse-monitoring and feature-specific storage controls apply until the service shuts down. The policy belongs to the provider route and account terms, so verify it again before production use.
Related comparisons
Head-to-head comparisons
Sora 2 vs Veo 3.1
Google's preview video model for text-, image-, and video-guided generation with native audio and output options up to 4K.
Open full comparisonSora 2 vs Veo 3.1 Fast
Google's lower-cost Veo 3.1 variant for quickly creating short video clips with native audio and output options up to 4K.
Open full comparisonSora 2 vs MiniMax H3 Max
fal's speed-focused post-trained version of MiniMax H3 for short video with native audio, fast 768p iteration, and text, image, or multimodal reference input.
Open full comparisonSora 2 vs Runway Gen-4.5
Runway's flagship developer video model for text- and image-guided clips, with professional formats and a predictable credit-per-second base rate.
Open full comparisonSora 2 vs Kling Video 3.0 Pro
Kuaishou's Pro video model delivered through fal for multi-shot text/image generation, optional native audio, references, and clips up to 15 seconds.
Open full comparisonSora 2 vs Kling Video 2.5 Turbo Pro
A cost-conscious Kling model on fal for fluid five- or ten-second clips generated from text or a starting image.
Open full comparisonSora 2 vs Grok Imagine Video 1.5
SpaceXAI's current video model for turning text, images, or references into short clips with optional generated speech and sound.
Open full comparisonSora 2 vs Seedance 2.5
ByteDance's current audiovisual model for coherent clips up to 30 seconds, with native sound, long-form storytelling, rich references, and targeted editing.
Open full comparisonSora 2 vs HappyHorse 1.1
Alibaba's unified video family for creating short clips with sound from text, a starting image, or a reference video.
Open full comparisonSora 2 vs LTX-2.5 Pro
Lightricks' quality-focused video model for making 720p or 1080p clips from text, images, or audio with native sound.
Open full comparisonSora 2 vs LTX-2.5 Fast
LTX's speed-optimized video model for text, image, or audio input, native sound, automatic duration, and output up to 4K.
Open full comparisonSora 2 vs LTX-2.3 Pro
LTX's mature full-workflow video model for generation plus Retake, Extend, Reframe, native audio, and first-to-last-frame control.
Open full comparison
