Short-term migration testing
Teams comparing an existing preview integration with the stable Omni 1.1 endpoint before switching production traffic.
Independent tool overview
Gemini Omni Flash Preview at a glance
Gemini Omni Flash Preview is Google's original paid Gemini API endpoint for short video generation and conversational editing. It remains callable at this review date but is deprecated and scheduled to shut down on September 30, 2026. New development should use gemini-omni-1.1-flash, and existing integrations should migrate now rather than treating this preview as a current product choice.
Visit the official Gemini Omni Flash Preview site ↗
- Model code
- gemini-omni-flash-preview
- Status
- Deprecated but still available at review
- Shutdown date
- September 30, 2026
- Replacement
- gemini-omni-1.1-flash
- Output
- 3-10 second 720p video with audio at 24 FPS
- Standard output cost
- Approximately $0.10 per second of 720p video
- Reviewed
- August 31, 2026
Overview
What Gemini Omni Flash Preview is
Google launched gemini-omni-flash-preview on June 30, 2026 as the first developer preview of Gemini Omni's video generation and natural-language editing. It accepts text, images, and short video, then returns a 3- to 10-second 720p clip with audio at 24 FPS through the Interactions API.
The preview demonstrated Omni's core interaction model: create a video, retain its interaction state, and request follow-up changes in plain language. Google allowed up to three sequential edits, but the launch version had important gaps: no scene extension, no uploaded audio references, unreliable processing of video references, and imperfect character consistency through scene changes or camera pans.
On August 27, Google released the generally available gemini-omni-1.1-flash replacement with scene extension, first-and-last-frame interpolation, and resolution controls from 360p to upscaled 4K. The preview endpoint is scheduled to shut down September 30, 2026, so this page is primarily a migration guide and historical record.
A model shutdown is an operational deadline, not merely an SEO label. Once Google turns the endpoint off, requests can fail and applications still pointing at it can lose their video feature. Teams should move traffic only after testing prompts, edits, safety handling, output delivery, cost controls, and generated-media review on the stable replacement.
Use cases
Who Gemini Omni Flash Preview is best for
The strongest fit depends on the job you need the product to complete, not the size of its feature list.
Reproducing a legacy result
Developers who temporarily need the original preview model to compare prompt or edit behavior during a controlled migration.
Historical product research
Readers documenting how Google's first developer-facing Omni video endpoint worked and which capabilities arrived in the stable release.
Capabilities
Core Gemini Omni Flash Preview features
Text-to-video
Generates a short 720p video with audio from a prompt describing the subject, action, setting, camera, style, dialogue, and sound.
Image-to-video
Uses an image plus text instructions to animate a concept or reference into a short clip.
Conversational editing
Allows follow-up natural-language revisions through the Interactions API while retaining the earlier generated clip as context.
Multimodal inputs
Accepts text, image, and video inputs, although the launch version's short video-reference path was documented as unreliable.
Native audio
Produces synchronized audio as part of the generated video rather than requiring a separate audio-generation request.
SynthID watermarking
Google says generated Omni videos include an imperceptible SynthID watermark for machine-detectable provenance.
Process
How the Gemini Omni Flash Preview workflow works
Step 1
Inventory every preview call
Search source code, configuration, serverless functions, queues, scheduled jobs, saved workflows, experiments, and fallback paths for gemini-omni-flash-preview. Record owners and traffic before changing anything.
Step 2
Move to the stable model code
Change new and test requests to gemini-omni-1.1-flash. Do not point new product work at an endpoint scheduled for shutdown.
Step 3
Review API differences
Compare the current Omni guide and SDK types with the preview integration, including response format, interaction state, URI delivery, resolution controls, interpolation, extension, regional restrictions, and unsupported parameters.
Step 4
Run representative comparisons
Test real prompts and source assets for visual quality, identity consistency, audio, safety blocks, latency, token usage, file delivery, and sequential edits. Keep the preview only as a temporary comparison baseline.
Step 5
Protect cost and content
Use server-side keys, budgets, rate limits, bounded retries, input rights and consent checks, least retention, and a human review of factual claims, people, brands, text, audio, and disclosure before publication.
Step 6
Cut over with a rollback window
Move a small cohort, watch failures and spend, then increase traffic. Roll back only within the remaining preview window and fix the stable path rather than relying on the deprecated endpoint long term.
Step 7
Remove the dependency
Delete preview configuration and fallbacks, update runbooks and tests, and alert on any remaining preview request before September 30, 2026. Verify that production no longer calls the retiring model.
Cost
Gemini Omni Flash Preview pricing and free plan
The preview has no free Gemini API tier. Its current Standard price matches Gemini Omni 1.1 Flash: $1.50 per million input tokens, $9 per million text output tokens including thinking, and $17.50 per million video output tokens. Google bills 720p output at 5,792 tokens per second, approximately $0.10 per generated second. Equal pricing is not a reason to keep the preview because its endpoint has a scheduled shutdown.
Free tier
Not available
The preview model requires paid Gemini API access.
- Google AI Studio availability does not create a production free tier for this model
- Consumer Google AI subscriptions are separate from Gemini Developer API billing
- Do not begin new implementation on the retiring endpoint
Standard paid API
$1.50 input / $9 text output / $17.50 video output per 1M tokens
Usage-based preview access until Google shuts down the endpoint.
- 720p video is billed at 5,792 output tokens per second
- Effective video-output cost is approximately $0.10 per second
- A 10-second output is roughly $1 before input, text output, retries, storage, and application costs
- Paid-tier prompts and responses are listed as not used to improve Google's products
Stable replacement
Same current Standard token rates
Gemini Omni 1.1 Flash is the recommended endpoint and adds production-oriented capabilities.
- Model code: gemini-omni-1.1-flash
- Generally available since August 27, 2026
- Adds extension, interpolation, and output-resolution control
- No shutdown date announced at review
Pricing checked . Check current pricing at the source ↗
Assessment
Gemini Omni Flash Preview strengths and limitations
Where it stands out
- The preview introduced a natural-language generation and editing loop in one multimodal Gemini model.
- Text, image, and short video inputs enabled more controlled concepts than text-only generation.
- Native audio created a more complete short prototype without a separate soundtrack model.
- A published replacement and shutdown date give existing users a concrete migration path.
- The preview and stable endpoint share the same listed Standard token rates at review, reducing pricing surprise during migration.
- SynthID provides a machine-detectable provenance signal for generated output.
What to consider
- The endpoint is deprecated and scheduled to shut down September 30, 2026. It is unsuitable for new production work and will become unavailable after retirement.
- The preview is limited to 720p 24 FPS output; the replacement adds 360p, 1080p, and 4K options, with 1080p and 4K produced through upscaling.
- The June preview did not support scene extension or uploaded audio references. Those capabilities should not be implied from the newer stable documentation.
- Google said video references up to three seconds were accepted by the preview API schema but not correctly processed by the model at launch.
- Google documented character-consistency limitations when changing scenes or using panning movement.
- Only up to three sequential conversational edits were described for the preview launch workflow.
- The API has no free tier, and repeated generations or edits can multiply usage cost quickly.
- Generated people, dialogue, text, actions, physics, products, and events can be inconsistent, inaccurate, or misleading even when the output looks polished.
- SynthID does not prove consent, rights, factual accuracy, or authorization. Teams still need human-readable disclosure where appropriate and a documented rights review.
- Do not use unverified output as evidence or as a substitute for licensed footage in news, medical, legal, financial, political, biometric, or safety-critical settings.
Compare
Gemini Omni Flash Preview alternatives
The right alternative depends on the specific output, workflow, controls and budget your project requires.
Miscellaneous
Gemini Omni 1.1 Flash
This is Google's direct stable replacement and the correct choice for new development or migration before the preview shutdown.
Explore Gemini Omni 1.1 Flash →Content Creator
Veo 3.1
Choose Veo 3.1 when Google's dedicated cinematic video model family fits better than Omni's conversational edit loop.
Explore Veo 3.1 →Content Creator
Veo 3.1 Lite
Choose Veo 3.1 Lite when lower-cost Google video generation is the priority.
Explore Veo 3.1 Lite →Content Creator
Runway Gen-4.5
Choose Runway Gen-4.5 when Runway's models and broader production interface are preferable to a direct Gemini API integration.
Explore Runway Gen-4.5 →Content Creator
Sora 2
Choose Sora 2 when OpenAI's video ecosystem and account or API model better match the project.
Explore Sora 2 →Questions
Gemini Omni Flash Preview FAQs
Is Gemini Omni Flash Preview still active?
Yes, as of August 31, 2026, but it is deprecated. Google lists September 30, 2026 as its shutdown date, so remaining use should be limited to migration and comparison work.
What replaces gemini-omni-flash-preview?
Google recommends gemini-omni-1.1-flash, the generally available endpoint released August 27, 2026.
When will Gemini Omni Flash Preview shut down?
Google's deprecation table lists September 30, 2026. Treat that as the deadline to remove production calls and fallbacks.
How much did the preview cost?
There is no free API tier. Standard pricing is $1.50 per million input tokens, $9 per million text output tokens, and $17.50 per million video output tokens, or about $0.10 per second of 720p video output.
What did the preview generate?
It generated 3- to 10-second 720p clips with audio at 24 FPS from text or images and supported conversational edits through the Interactions API.
Did the preview support video extension?
No. Google's June launch post said scene extension was not yet supported in the Gemini API for the preview. Extension arrived with Gemini Omni 1.1 Flash.
Can I just change the model string?
The model code change is necessary but should be followed by representative tests of prompts, sequential edits, parameters, output delivery, safety blocks, costs, resolution, storage, and user-visible quality before production cutover.
Does the preview add a watermark?
Google says Omni output includes invisible SynthID watermarking. Keep provenance attached and add visible disclosure when the audience, platform, law, or context requires it.
Bottom line
Our Gemini Omni Flash Preview verdict
Gemini Omni Flash Preview was a useful first look at Google's conversational video API, but it is now a migration dependency rather than a tool to adopt. Its 720p generation, native audio, and follow-up editing remain usable briefly, while the stable 1.1 release adds meaningful capabilities at the same listed Standard rates. Inventory every call, validate representative work on gemini-omni-1.1-flash, move production traffic, and remove the preview path before September 30, 2026.
Visit Gemini Omni Flash Preview website ↗Similar Content Creator Tools
View all Content Creator AI tools →
OpenArt Director
OpenArt Director - Vibe-directing tool that builds five-minute films from chat
.gif)
Nano Banana 2 Lite
Nano Banana 2 Lite - Google's high-volume, cost-effective image model that generates pictures in just four seconds
Palmier
Palmier - Generate, edit, and export production-ready AI videos without leaving your timeline
Seedance 2.0
Seedance 2.0 - ByteDance's frontier AI video model
AI training for the future of work.
Get access to all our AI courses, hundreds of real-world AI use cases, live expert-led workshops, an exclusive network of AI early adopters, and more.
AI Courses
Get unlimited access to all of our current & upcoming industry-specific AI courses for the duration of your subscription.
Daily Guides
To keep up with the rapid pace of AI, our team publishes AI implementation guides daily. Our library contains 300+ practical use cases to automate real-world work.
Workshops
Join weekly, live, interactive sessions with industry leaders who are at the forefront of AI for hands-on implementation guidance and exclusive insights.
Community
Network with an exclusive community of AI-first professionals who are working smarter with AI. Learn how early adopters are using AI in their work and businesses.