Short answer
For a new video-generation API workflow, Gemini Omni 1.2 Flash is the available option of these two; Sora is discontinued. OpenAI says the Sora web and app experiences ended on April 26, 2026, and the Sora API ended on September 24, 2026. Google says Gemini Omni 1.2 Flash is generally available on Google Cloud. Its documentation lists support for the Interactions API. That establishes an availability and API-support distinction—not proof that Gemini produces better video or a separate “production-ready” designation. [1][2][3]
What “Gemini vs. Sora” means
The comparison needs a specific model and channel. Here, Sora means OpenAI’s discontinued video product and API; Gemini means Google’s Gemini Omni 1.2 Flash video-generation model on Google Cloud’s Gemini Enterprise Agent Platform. Google’s model documentation lists video generation and editing capabilities and support for the Interactions API. This is not a comparison of the Gemini chat app with Sora, nor a like-for-like quality benchmark. [1][2][3]
| Question | Sora | Gemini Omni 1.2 Flash |
|---|---|---|
| Can you start a new video workflow now? | No. OpenAI’s Sora web/app service and API have been discontinued. | Google says the model is generally available on Google Cloud; its documentation lists API support. |
| Video generation | No current Sora service to use. | Google documents video generation and editing with synchronized audio, up to 15 seconds, at 360p, 720p, 1080p, or 4K. |
| API status | No; the Sora API shutdown date was September 24, 2026. | Google documents Interactions API support. The retrieved wording establishes general availability, not a separate “production-ready” designation. |
| Published video price | Not applicable to a discontinued API. | Google announced $0.07 per second of 720p video output. Other resolutions may have different rates. |
[1][2][3]
Gemini Omni 1.2 Flash: documented capabilities and trade-offs
Google documents Gemini Omni 1.2 Flash for video generation and editing. The Agent Platform video guide specifies clips from 3 to 15 seconds at 360p, 720p, 1080p, or 4K; the model page lists synchronized-audio video output. Google’s launch announcement says audio references require allowlisting. So distinguish audio included in generated output from access to audio-reference input: the latter is gated. [2][3][4]
These features make Omni a candidate for short marketing clips, product visuals, and iterative creative work. They do not establish that it will outperform Sora’s former model. For output quality, test your own prompts and assets for adherence, continuity, audio needs, editing effort, and cost per accepted result; this article does not claim to have run that comparison.
Production API access and price
Gemini Omni 1.2 Flash is available through Google Cloud’s Gemini Enterprise Agent Platform API using the Interactions API. Google’s reference documents the create-interaction route at https://aiplatform.googleapis.com/v1/projects/{project}/locations/global/interactions. Google’s setup guide recommends Application Default Credentials for production and describes API keys as a testing option. The model page notes that deploying its example app requires a Google Cloud project with billing and the Agent Platform API enabled. [3][5][6]
Google says Gemini Omni 1.2 Flash is generally available on Google Cloud, and its documentation lists Interactions API support. The evidence cited here does not establish a separate “production-ready” label. General availability does not guarantee access in every account or location: Google documents regional and global endpoints, so check the supported-model locations and your project’s configuration before relying on a particular endpoint. [2][3][7]
Google’s published price is $0.07 per second of 720p video output, in USD. At that rate, a 10-second 720p output is $0.70, and a 15-second output is $1.05. These are arithmetic examples based on the published rate, not measured bills. Google says other resolutions have similar savings but does not give their exact rates in the announcement excerpt cited here. Confirm current model-specific pricing and billing details before forecasting spend. [2]
There is no current API price comparison with Sora: its API has shut down. Historical Sora pricing would not tell you the cost of running a new Sora production workflow today. [1]
Which should you choose?
- Building a new video-generation API workflow: Evaluate Gemini Omni 1.2 Flash on Agent Platform. Sora is discontinued.
- Maintaining an existing Sora integration: Plan around the API shutdown. Identify where video requests and retries occur, then test a replacement before migrating production traffic.
- Choosing on output quality: Don’t rely on old comparison articles or assume a universal winner. Test the same brief and source assets, then assess usable clips, prompt adherence, continuity, audio, editing effort, and cost per accepted output.
- Using Gemini for planning rather than video output: Keep that separate from the model/API comparison. A text assistant can help draft a script or shot list, but that alone is not a video-generation API.
Bottom line: For a new video-generation API workflow between Gemini and Sora, Gemini Omni 1.2 Flash is the practical option—not because a head-to-head test proves it is better, but because OpenAI has discontinued Sora and Google says Omni 1.2 Flash is generally available, with Interactions API support documented. Check project setup and supported locations before deployment. [1][2][3][7]
Sources
What to know about the Sora discontinuation
Gemini Omni 1.2 Flash now GA: cost-effective video
Generate videos with Agent Platform