2026-11-07 · VideoFreeTier

Is Gemini Omni Flash free: the Flow free plan stops at 10 seconds and the API free tier prints Not supported

There is no single yes or no, because Google publishes the answer on three separate product surfaces and none of them says what the other two say. On 7 November 2026, the Google Flow price table lists Omni 1.2 Flash (≤10s) inside the free plan; the Gemini API's own rate-limit table prints Not supported in both cells of its Free row for gemini-omni-1.2-flash; and the Gemini app subscription page's free column lists no video-generation feature at all. The first is a browser plan with a duration ceiling, the second is a usage tier on a developer API, and the third is a feature list that cannot be read as a capability switch. Treat each as a statement about its own channel only.

Everything below was read from public pages on 7 November 2026 without registering, paying, or making an API call. Where a figure belongs to a different product or a different model version, it is labelled. Where an official page stays silent, this page says so rather than filling the gap.

Three pages, three different questions

A search for is gemini omni flash free is really several questions wearing one coat: can I generate a clip in a browser without a subscription, can I call the model from code on the free API tier, and does the Gemini app's free tier include video generation at all. Google answers the first on flow.google.com/about, the second on the Gemini API rate limits and pricing pages, and the third — or rather, declines to answer it — on gemini.google/us/subscriptions/. The reader who opens only one of them gets an answer that sounds complete and is not.

None of these rows is a verdict on the model. A plan table describes what a plan grants, a rate-limit table describes what a usage tier permits, and a feature list describes what a page chose to enumerate. Writing that the model is not free from the API row, or that it is free from the Flow row, converts a channel fact into a model fact — the most common error in answers to this question. This page quotes the en rendering of both consumer pages, because a US reader following the links sees English text. flow.google.com/about?hl=en is what is quoted below, and gemini.google/subscriptions/?hl=en resolves to gemini.google/us/subscriptions/?hl=en with the plan selector set to United States. Reading those same two pages in zh-CN and ko-KR earlier on the same date produced the same Flow prices and the same feature lines, but a different free-tier price string on the subscription page. Locale is a real variable on that surface, not a cosmetic one.

Door one: Flow's free plan names Omni, with a duration ceiling

The Flow price table's first card is labelled Without a Google AI Subscription / Free of Charge. Its Included Features: list opens with 50 daily Google Flow credits, and the second line is Omni 1.2 Flash (≤10s). That parenthetical is the whole answer for the browser path: the free plan names this model, and the name it prints is capped at ten seconds. The same free card also lists Nano Banana Pro, Nano Banana 2.1, Veo 3.1 Lite, Google Flow Tools (Usage only), Agent, Text to Video, Frames to Video, Ingredients to Video, Video-to-video editing, Scenebuilder, Characters, Avatars, and 2K image upscaling. Video creation is therefore not absent from the free card; what is absent is any video upscaling line.

The next card, Google AI Plus at $7.99/mo, adds Omni 1.2 Flash (11-15s), Video Extension (with Ingredients), Google Flow Tools (Creation), 720 & 1080p video upscaling, and Higher image generation limits, on top of Everything for free users. Google AI Pro at $19.99/mo adds AI credits available (top-ups). The 4K image and video upscaling line sits on the $99.99/mo Ultra card, and a second Ultra card at $199.99/mo prints Everything in Ultra $99.99 with 25,000 monthly Google Flow credits. Each price string carries the table's footnote asterisk, kept here as a footnote marker rather than a printed glyph so the site's renderer does not read it as emphasis; the footnote itself reads Prices may vary by market. The page also states Features may vary by Google AI subscription tier, platform (web vs. mobile), and region. 18+. All of these lines are the en rendering of flow.google.com/about?hl=en read on 7 November 2026.

One price conflict has to stay visible. The Flow price table prints Google AI Plus at $7.99/mo, while the US subscription page prints Google AI Plus at $4.99/ month on the same date. Both are official Google pages, both name the same plan, and neither explains the difference; this page quotes both figures as printed and neither reconciles them nor picks one as the true price. The subscription page also states Google AI Plus is available in more than 160 countries and territories, which is consistent with market-dependent pricing but does not resolve the two numbers. Two boundary conditions are easy to miss. First, 50 daily Google Flow credits appears on the paid cards too; the paid tiers add monthly credits rather than replacing the daily line, so the daily 50 is not a free-tier exclusive. Second, the page's FAQ asks How much does it cost to generate images and videos? but no answer appears in the captured text, and no line on the price table states a per-generation credit cost for Omni. Correction, 14 November 2026: that last sentence is still true of the price table, but the conclusion this page drew from it on 7 November 2026 was wrong. A fourth official page, the Flow help centre's Manage your Google Flow credits (support.google.com/flow/answer/16526234?hl=en), does publish the rate: Generate or edit videos up to 10s for All users costs 1 credit per second at 360p and 3 credits per second at 720p, and the companion models page caps the free plan at 3 to 10s video length. The price table does not link to that help page, which is why reading only the price table makes the figure look unpublished. The daily 50 can now be divided by a published rate; Google Flow free credits carries that arithmetic and this page does not repeat it. The free card also publishes no output resolution for Omni: 2K image upscaling is about still images, while 720 & 1080p video upscaling is on Plus.

Door two: the API free tier is literally marked Not supported

The Gemini API rate-limits page has a dedicated Gemini Omni Flash rate limits section whose table reads, verbatim: Free / Not supported / Not supported for requests per minute and requests per day; Tier 1 / 8 / 40; Tier 2 / 16 / 200; Tier 3 / 40 / 2,000. The page defines Free as Active project or free trial; some projects need verification, and states that moving from Free to a paid tier requires setting up billing in AI Studio. Read on 7 November 2026, that is a statement about today's Free usage tier for gemini-omni-1.2-flash — not a promise that no free tier will ever exist, and not a statement that the model itself refuses free use on every product.

The pricing page says the same thing in prose and should be quoted rather than paraphrased: Our next-generation video generation and editing model supporting up to 15 seconds of video generation and editing with native audio, now generally available to developers on the paid tier of the Gemini API. What is paid is the tier. Turning that line into a claim that the API itself costs money to exist loses the distinction the sentence is making. The same page's billing footnote gives the standard rate: 5,792 tokens per second of 720p video, Under Standard pricing ($12.00 per 1M tokens), approximately $0.07 per second for 720p, with $0.023/s at 360p, $0.104/s at 1080p with upsampling and $0.209/s at 4K with upsampling. Those are the page's own figures, not this site's arithmetic.

One inconsistency is worth keeping visible. The 1.1 Flash section prints the same 5,792 tokens per second of 720p video line but concludes approximately $0.10 per second without naming a token price, while 1.2 names $12.00 per 1M tokens and concludes $0.07. The pricing page does not explain the difference between the two effective rates, and this page neither reconciles them nor picks one as the truth. Both lines are quoted as printed on 7 November 2026; neither is used here to compute a free-tier equivalent.

Door three: the Gemini app's free column is silent, not prohibitive

The gemini.google/us/subscriptions/?hl=en page, read with its plan selector set to United States on 7 November 2026, prices the free tier at $0/ month with a Google Account. Its free-column feature list runs Gemini app, Access to our Flash-Lite model, Image generation and editing, Deep Research, Gemini Live, Canvas, Gems, Google Flow, Gemini Notebook, and 15 GB of cloud storage across Gmail, Drive, and Photos. The Google Flow entry reads Access to our AI creative studio to create cinematic scenes and stories, including limited access to Nano Banana Pro. There is no separate Video Generation item in that free column; the occurrences of that phrase elsewhere on the page are navigation labels, not entries in the free feature list.

Silence is not a denial. The Plus and Pro descriptions on the same page use comparative wording about higher usage limits followed by access to more features like video generation and Daily Brief — more features such as video generation. That is marketing language about limits, not a switch that turns a capability on. This page therefore does not conclude that the free Gemini app cannot generate video; it concludes only that the free feature list does not enumerate it, and that the only per-model duration ceiling published for a free channel is the one on the Flow price table. A table that omits a line is not a statement that the line is false.

The same subscription page also ties Flow credits directly to this model: its paid rows say Get 200 Google Flow Credits to use across our AI creative studio to create cinematic scenes and stories with access to Gemini Omni Flash and custom tool creation. That is a second official page naming the model behind the Flow entries, consistent with the Flow price table's Omni 1.2 Flash labels. It still publishes no per-generation price in credits.

What changes on 3 December 2026

The older 1.1 model is the obvious workaround to look for, and the pricing page closes it with an explicit date: Effective December 3, 2026, all API requests to gemini-omni-1.1-flash automatically route to gemini-omni-1.2-flash and are billed at gemini-omni-1.2-flash rates. The model page repeats it in its version list: Stable: gemini-omni-1.1-flash (routes to gemini-omni-1.2-flash on December 3, 2026), alongside Shut down: gemini-omni-flash-preview. The model page also separates the two generations' output ranges — 1.2 prints Output video 3s-15s (360p/720p/1080p/4K, 24 FPS; 3s-10s for first and last frame interpolation and per-turn video extension up to 30s total), while 1.1 prints Output video 3s-10s (360p/720p/1080p/4K, 24 FPS).

So the wait-for-the-older-version exit closes on that date for API callers, because requests are rerouted and billed at 1.2 rates. The 1.2 range on the model page is an API specification; it is not a promise about what the Flow free plan's ≤10s card will output, and the two should not be read into each other. Both readings are dated 7 November 2026.

What people are actually asking

Autocomplete suggestions collected from Google's own suggestion endpoint on 7 November 2026 include is gemini omni flash free, gemini omni flash free tier, gemini omni flash api free, gemini omni flash free trial, gemini omni flash free use, and how to use gemini omni flash for free. Suggestions show which phrasings exist; they are not evidence about the answer, and this page quotes no ranking, volume or difficulty figures. Comment threads around videos about the model show the same split in plainer language: one under 8 Insane Ways to Use Gemini Omni Flash asks, This might be a dumb question BUT if you are using Gemini Omniflash, can’t you just do it directly in the Gemini App? In other words, what is the added value of ElevenLabs in this specific scenario? — the app-versus-tool confusion the third door addresses. Another asks What’s the workflow for videos longer than 10 seconds? — exactly the boundary printed on the Flow free card. Both are quoted as reader questions, not as evidence about product behaviour.

Which door to use

If you want a browser workflow with no subscription, the free Flow plan is the channel that names Omni 1.2 Flash (≤10s); read that as a plan-table grant with a ten-second label, not as a resolution guarantee, because the price table publishes no free output resolution. If you need the 11–15 second band, Video Extension (with Ingredients), or 720 & 1080p video upscaling, those lines sit on the Google AI Plus card — which the Flow table prints at $7.99/mo and the US subscription page prints at $4.99/ month, two figures this page does not choose between — and 4K upscaling sits on Ultra. If you are calling the model from code, the API Free row reads Not supported / Not supported on 7 November 2026; the paid-tier standard rate printed on the pricing page is approximately $0.07 per second at 720p. And if your question is only about the Gemini app, the honest answer is that its free feature list does not say.

Prices, tiers and model availability change without notice, and the Flow table itself warns that prices may vary by market. Every reading on this page is dated 7 November 2026 with the channel named. Re-read the source before quoting any of it.

Correction, 14 November 2026: three rules behind the daily 50

Manage your Google Flow credits (support.google.com/flow/answer/16526234?hl=en), read in en on 8 November and re-checked on 14 November 2026, prints three rules the price table never links to. Your daily credit refresh is triggered by your first generation. If you don't use the full 50 Google Flow credits in a day, the unused Google Flow credits don't roll over. During peak hours, roughly 2 PM to 5 PM UTC, you may not be able to generate videos without a subscription. That line limits non-subscribers in a window, not the model. Rates are per generation, not per request; Google Flow free credits carries the clip counts. No generation was run here, so these rules are quoted, not measured.

What this page does not settle

No account was created, no subscription purchased, no credit balance observed, and no API call made, so nothing here measures a generation: not a real debit for one Omni clip, not the free plan's actual output resolution, not the visual result of 4K image and video upscaling, and not whether the free Gemini app produces video in practice. The Flow price table's own FAQ heading asks what image and video generation costs, but the captured text carries no answer; that table is silent, but Google does publish per-generation credit rates on the help page quoted above. This site did not run a generation, so those rates are quoted from an official page rather than measured in an account; an actual debit remains 本站未实测. Google's own search results page returned a JavaScript shell on this machine, so this page makes no claim about which results rank for the question. Both consumer pages were re-read in en and en-US on 7 November 2026 with the subscription page's region selector on United States; the Flow table warns that features may vary by tier, platform and region and that prices may vary by market, so a reader in another market should not expect either the wording or the numbers to match. Nothing here is a measurement of a free tier in use.

Related on this site: Veo 3.1 free tier — what a version number does to a free-tier answer. Seedance 2.5 API free — another case where a free-token headline and a model's own table row disagree. Kling 4.0 API pricing — a different API's unit-based meter. The observed figures behind the model pages live in the data page.