How Much Does Gemini Omni Cost?
Gemini Omni Flash charges $1.50 per million input tokens across text, image, video and audio, $9 per million text output tokens, and $17.50 per million video output tokens. Google puts the video rate at roughly $0.10 per second of 720p output.
Google DeepMind announced the Gemini Omni family at Google I/O on May 19, 2026, as an any-to-any multimodal model: text, image, audio and video in, finished video out. Omni Flash was the first version to ship.
API prices were not public on announcement day, which is why launch coverage and the per-token rates below answer two different questions. Here are the rates as they stand on Google's official Gemini API pricing page:
| Line item | Price per 1M tokens |
|---|---|
| Input (text, image, video, audio) | $1.50 |
| Text output | $9.00 |
| Video output | $17.50 |
Video output is the row that catches teams out, and it is covered in full further down. AI Perks tracks the credit programs that offset a bill like this.
Why Gemini Omni Bills Video By the Token, Not By the Second
Google meters Omni video the same way it meters text: by output token. The familiar per-second figure is a conversion, not the underlying unit.
The published rate is $17.50 per million video output tokens. Google's own equivalence for that is roughly $0.10 per second of 720p video, and every per-second number here comes from that approximation, not from a verified token count.
Two practical consequences follow.
Your invoice will be denominated in tokens, not seconds. A team that forecasts in seconds and reconciles in tokens ends up with a budget that does not match the bill, and the gap only shows up after the spend.
The per-second figure is tied to 720p. A precise tokens-per-second conversion, and how it moves at other resolutions, is not something we could confirm. Plan at the 720p figure and treat every other resolution as unknown.
Using Google's own per-second approximation, the planning arithmetic looks like this:
| Output at 720p | Approximate cost |
|---|---|
| One 8 second clip | ~$0.80 |
| One 10 second clip | ~$1.00 |
| One 30 second clip | ~$3.00 |
| One minute of finished video | ~$6.00 |
| 100 ten second clips | ~$100 |
| 1,000 ten second clips | ~$1,000 |
The rates are Google's, the multiplication is ours, and it rounds. Text output bills separately at $9 per million, so a workload returning narration or captions alongside video carries both meters.
Is Gemini Omni Free?
Yes, on Google's consumer surfaces: Omni is bundled into the Gemini app and shipped as a free route inside YouTube Shorts and the Create app. What that does not settle is the developer side, because we could not confirm any published free-tier quota for Omni.
The routes documented as free to a person rather than to running code:
- The Gemini app. Omni is bundled into Google's consumer plans, spanning the free tier through AI Plus at $4.99, Pro at $19.99 and Ultra at $100.
- YouTube Shorts and the Create app. Reported at launch as a free consumer route. Google has not published usage limits or generation caps for it, so treat any specific quota you read elsewhere as unconfirmed.
Elsewhere in the lineup the picture is simpler. Gemini 3.6 Flash, Gemini 3.7 Flash and Gemini 3.5 Flash-Lite are all listed with a free tier, and Nano Banana 2 is free to use through the Gemini app and Google AI Studio.
So the honest answer to "is Gemini Omni free" is: yes for making clips by hand, unconfirmed for anything called from code at volume. Nobody has published where the allowance ends, so budget as though it ends early. That gap is what credits are for, and it is why AI Perks exists.
Gemini Omni vs the Rest of the 2026 Gemini Lineup
Omni is the expensive specialist in a lineup that otherwise got cheaper through 2026. Its $1.50 input rate is double the Gemini 3.6 Flash introductory rate and five times Gemini 3.5 Flash-Lite.
| Model | Input per 1M | Output per 1M | Documented free access |
|---|---|---|---|
| Gemini Omni Flash | $1.50 | $9 text, $17.50 video | Gemini app, YouTube routes |
| Nano Banana 2 (Gemini 3.1 Flash Image) | $0.50 | $60 images | Gemini app, AI Studio |
| Gemini 3.6 Flash | $0.75 | $3.75 | Free tier |
| Gemini 3.7 Flash | $0.75 (reported) | $3.75 (reported) | Free tier |
| Gemini 3.5 Flash-Lite | $0.30 | $2.50 | Free tier |
| Gemini 3.5 Flash Cyber | Not published | Not published | None |
Three footnotes decide whether this table is still accurate when you read it.
The Flash rates are introductory. Gemini 3.6 Flash is priced at $0.75 and $3.75 through December 31, 2026, with a standard rate of $1.50 and $7.50 after that. That is a published doubling, not a rumour. Gemini 3.7 Flash is reported at the same rates through the same date, but from secondary coverage rather than a page we could confirm, and its post-introductory rate is not on the record.
Gemini 3.5 Flash Cyber has no public price because it is not a public product. It is a limited-access cybersecurity model, built on 3.5 Flash and used inside Google's CodeMender agent, offered to governments and trusted partners. Anyone quoting a per-token rate for it is guessing.
Nano Banana 2's $60 output rate is not what it looks like. Google's pricing page puts it at roughly $0.045 to $0.151 per image depending on resolution, 0.5K to 4K, with Batch pricing at 50% off. Google has not confirmed an exact launch date, though a Nano Banana 2 Lite variant appears in its own blog listing under June 2026.
What Google Has Not Published About Gemini Omni
Four things are still missing from the public record, and every one of them is somewhere a rushed article invented a number.
The exact token conversion, and per-resolution rates. Google frames the video price as roughly $0.10 per second at 720p. A precise tokens-per-second figure, and rates for 360p, 1080p or 4K output: the vendor has not published this.
Free-tier quotas. Google bundles Omni into the Gemini app and the YouTube Create route without publishing per-plan generation caps, and we could not confirm a developer free-tier allowance for it either. Any hard number for daily clips is someone's estimate.
A Batch rate for video output. Google lists a 50% Batch discount for Nano Banana 2. We could not confirm an equivalent published Batch rate for Omni video output, so forecast at the full $17.50.
Any price for Gemini 3.5 Flash Cyber. Not withheld, not delayed, simply not a public product.
Writing "the vendor has not published this" is less satisfying than a number. It is also the only version of this page that survives to December.
Why Video Output Is the Category Where Credits Matter Most
Video is the most expensive output category in the Gemini lineup, which makes it the category where cloud credits carry the most value. At roughly $0.10 per second, a single grant covers a great deal of finished video.
The two categories worth filtering to at getaiperks.com are AI Model and Cloud Infrastructure, which cover the model calls and the compute underneath them.
Model the token bill before the build, not after. Take expected seconds of 720p output at the published per-second figure, then add the text meter running alongside it at $9 per million. If the number is uncomfortable, change the architecture rather than the budget: shorter clips, fewer regenerations and a cheaper model for the non-video half of the pipeline all move it further than any rate will.
One date is worth a calendar entry. Introductory Flash pricing runs through December 31, 2026, and credit terms move at least as often as model prices do.
AI Perks covers $7.7M in credits across 194 companies. For Gemini specifically, the route that comes up is Google's own startup cloud program.
Frequently Asked Questions
How much does Gemini Omni cost per second of video?
About $0.10 per second of 720p output. Google bills video by output token at $17.50 per million and puts the equivalence at roughly ten cents a second, so a ten second clip lands near $1.00. The precise conversion is not published, so treat any per-second figure as an approximation. Credits that offset Gemini API spend are listed at getaiperks.com.
Does Gemini Omni have a free tier?
Yes, on the consumer side. Omni is bundled into the Gemini app across the free plan and the paid ones, and launched as a free route inside YouTube Shorts and the Create app. We could not confirm published quotas or generation caps for Omni, so where the free allowance ends for a developer is not on the record.
When did Gemini Omni launch?
Google DeepMind announced the Gemini Omni family at Google I/O on May 19, 2026, with Omni Flash as the first shipping version. API pricing was not public at the announcement, so the per-token rates arrived later than the model did.
Is Gemini Omni more expensive than the other Gemini models?
Yes, by a clear margin. Input is $1.50 per million against $0.75 for Gemini 3.6 Flash and $0.30 for Gemini 3.5 Flash-Lite, and video output at $17.50 has no equivalent elsewhere in the lineup. AI Perks tracks the credits that close that gap.
What does Gemini 3.5 Flash Cyber cost?
Google has not published a price. Gemini 3.5 Flash Cyber is a limited-access cybersecurity model for governments and trusted partners, used inside Google's CodeMender agent, rather than a general API product. No public per-token rate exists, so treat any quoted figure as unconfirmed.
Can I get free Google AI credits for video generation?
Google runs a startup cloud credit program, and several other providers run comparable programs for video and compute. Amounts and terms vary and change often, so trust the figures on the program's own page. AI Perks tracks $7.7M in credits across 194 companies in one place.
With Gemini Omni the deliverable is finished video and the meter is a token counter. Get the conversion right and budgeting is arithmetic. Get it wrong and the invoice arrives in a unit nobody forecast.
Generate the video. Let someone else pay for the tokens.