Gemini Omni 1.1 in Google Vids: Free 1080p Is a Clip Desk, Not a Script

Gemini Omni 1.1 in Google Vids: Free 1080p Is a Clip Desk, Not a Script

Free HD is a generate desk. It is not a signed line. On 23 September 2026, Google product manager Justin Luk wrote that Google put Gemini Omni 1.1 Flash into Google Vids so anyone with a Google or Google Workspace account can generate high-quality video at no cost. The post is about 1080p scenes, duration control, scene extension, templates, and a SynthID watermark. It does not ship a legal review, a pack-shot lock, or the voiceover tool — that last one is labeled coming soon. ClipCanva is not affiliated with Google, VEED, Kapwing, Synthesia, Runway, or Canva. Lock the claim on the AI script generator, lock the still on image to video if the SKU already exists, then generate motion on the AI video generator. Do not paste an unfinished paragraph into vids.new and call take one a launch.

Launch facts (confirmed 24 September 2026)

Source	Public claim	Workflow implication
Google: Gemini Omni in Google Vids (23 Sep 2026, Justin Luk)	Newest models in Google Vids. Anyone with a Google or Google Workspace account can generate high-quality videos at no cost using Gemini Omni 1.1 Flash and new creative controls.	“At no cost” is an access sentence. It is not a quota, a max duration, or a license to skip the script.
Same Google post	Desktop path: visit vids.new and select Create AI videos.	This is a browser generate desk. Do not brief it as a phone app or as Flow.
Same Google post	Extend scenes while keeping visual context, lighting, character appearance, and environment consistent. Set super-specific durations so the clip aligns with story and voiceover. Generate in HD: brand-new AI scenes in full 1080p, or upscale existing AI clips on the timeline.	Duration control is a timing desk. It does not invent the VO. Google did not publish a max clip length for this Vids path. Do not copy Flow’s older numbers onto this button.
Same Google post	Use cases named in the post: ready-made templates for a product launch; photos of new arrivals turned into promo videos; landing pages, social announcements, and digital event signage; community events and fundraisers.	Photos-of-stock is image-to-video. A landing-page claim is script. Do not mix them in one prompt.
Same Google post	Every clip generated with Omni 1.1 in Google Vids includes an imperceptible SynthID watermark in the frames so viewers can verify the clip was generated with AI.	Watermark is provenance. It is not a brand-safety review of your claim.
Same Google post	Coming soon: Gemini 3.8 Flash-Lite text-to-speech in Google Vids will turn a script into voiceover across 100+ languages.	TTS is not claimed as live. Do not book a multilingual VO deadline on this sentence.
Same Google post	More generation on a personal account: Google AI plans. Teams: Workspace Business and Enterprise get expanded generation pools and admin controls.	Free is the floor. Capacity and admin are a different SKU. Google did not publish a free-generation count in this post. Do not invent one.
Google: Build with Gemini Omni 1.1 Flash (27 Aug 2026)	Omni 1.1 Flash for developers: scene extension, first and last frames, 360p drafts, 1080p/4K upscale, available in Google AI Studio and Flow for Google AI Plus/Pro/Ultra.	That is the Flow / API desk from August. Today’s Vids post is a Workspace browser desk at 1080p. Do not quote 4K or first/last-frame as Vids facts unless Google says so on the Vids page.
Google Workspace: Vids	Workspace product page: generate video from a prompt; customisable avatars; animate uploaded images; templates and Slides; export to YouTube (exports default to private).	Avatars, stock library, and YouTube export are editor jobs. They sit next to generate. They are not the 23 September Omni 1.1 announcement.

Google did not publish prices, a free-tier generation cap, or a maximum Omni clip length for Vids in the 23 September post. Confirm the live picker before you quote seconds, 4K, or a VO language list.

What Google Vids just added — and what it still is not

Surface	What is actually on the public record	What operators often assume	Fail condition
Omni 1.1 in Google Vids	Free-with-account 1080p generate, duration control, scene extension, SynthID	A finished campaign, including VO and legal	Shipping take one because it was free
vids.new → Create AI videos	Desktop generate entry	A mobile short-form studio	Building a field team’s workflow on a desktop-only path
Scene duration	Align the clip to a story/VO you already have	The model writes the VO to that length	Setting “8 seconds” on a paragraph that needs 20
Scene extension	Longer scene, same lighting/character/environment	A second shot, a new location, a new SKU	Extending until the bottle warps
1080p generate / upscale	New AI scenes at HD, or upscale AI clips already on the timeline	4K broadcast masters	Quoting Flow 4K on a Vids export
SynthID	Imperceptible watermark in generated frames	Platform approval, or “not AI”	Stripping or ignoring provenance on a regulated claim
Gemini 3.8 Flash-Lite TTS	Coming soon, 100+ languages	Live multilingual VO today	Booking a 12-language cutdown on a feature that is not out
Workspace Vids editor	Prompt generate, avatars, image animate, templates, private YouTube export	One-click from brief to paid media	Skipping the script because the editor looks complete

A free 1080p button is still a clip desk. If you do not know the one sentence the viewer must remember, you are burning a free quota on a draft you cannot defend.

Pick the desk before you open vids.new

The fail is treating “Google made HD free” as “Google wrote the ad.” Map the job first.

Job	Where it actually runs	What you must supply	What the system may invent	Fail condition
Legal line, spoken words, on-screen claim	AI script generator	Audience, one claim, duration, forbidden words	Extra scenes, unearned superlatives	Prompting Vids from a slogan you never signed
Shot list / clip-length beats	Script output + a scene card you keep; reusable language on prompt ideas	Shot size, camera move, cut, seconds	Decorative adjectives that fight the still	One paragraph dumped into Create AI videos
Still that must stay on-model	Stills desk, then image to video or Vids’ “animate your images” path	Approved pack shot as the first frame	Camera path, warped type, extra logos	Generating the bottle from a text prompt because HD is free
Finished campaign clip	AI video generator or a named model on Compare	Locked still and/or locked caption, crop, seconds	Extra props, a second spoken line	Shipping the first free Vids take because it “looks HD”
Long recording → short spoken plan	AI video summarizer, then script	Source cut, claims you will keep	Compression that changes meaning	Asking Omni to “just recap the webinar” with no card
Existing timeline footage you must restyle, not replace	Runway in DaVinci Resolve (23 Sep 2026): plugin for DaVinci Resolve Studio 19+; restyle the clip under the playhead with Aleph 2.0. Free plugin; generation uses paid Runway credits.	In/out, one change request	New props, identity drift	Regenerating the hero in Vids to avoid a round trip
Script + visuals + narration in one vendor button	VEED AI Video Generator: describe a concept or enter a script; output marketed as visuals plus realistic narration	A script you accept	Footage, VO, a second claim	Treating VEED’s one-shot as proof Google’s TTS is live

If the brief is “this bottle already exists; make it move,” that is image-to-video from a locked still. If the brief is “we do not know the shots yet,” that is script. If the brief is “HD and free, so generate now,” you do not have a brief.

Google Vids vs VEED vs Kapwing vs a restyle plugin

Dimension	Google Vids + Omni 1.1 (23 Sep)	VEED generate	Kapwing generate	Runway in DaVinci Resolve (23 Sep)
Official surface	vids.new → Create AI videos; Workspace Vids	VEED AI Video Generator	Kapwing AI Video Generator	Runway plugin for Resolve Studio 19+
What they sell on the page	Free-with-account 1080p Omni clips, duration, scene extension, SynthID	Text / script / image in; video with visuals, VO, avatars, subtitles	Free, online, no download; separate script generator and video generator in the nav	Stay in the NLE: generate or restyle under the playhead, import to the timeline
Script desk	You bring it. In-Vids TTS is coming soon	Script is an input they advertise	Script Generator is a separate tool from Video Generator	Not a writer. Restyles footage you already cut
Provenance	SynthID on Omni 1.1 Vids clips	Not claimed in the VEED hero we checked	Not claimed in the Kapwing hero we checked	Your original take remains on the timeline under the restyle
Do not use it for	Writing the claim, or quoting a max duration Google did not publish	Proof that Google’s 100+ language VO is live	A pack-shot lock	Inventing a product that was never shot

VEED’s public promise is one-shot completeness (script in, narration out). Kapwing’s public IA splits script and video into different tools — that split is the correct operator habit. Google’s 23 September post made generate cheap and HD. It did not merge the script desk into the generate desk.

Creator / operator checklist

1. Write the one line the clip must carry. Duration control is useless if the sentence does not fit. Use the AI script generator. Strike “#1,” ingredient claims, and competitor names you cannot defend.
2. Decide whether a still already exists. If the SKU is real, lock the pack shot, then image to video or Vids’ image-animate path. Do not ask Omni to invent the bottle because 1080p is free.
3. Open the right Google surface. vids.new on desktop, Create AI videos. Do not brief Flow first/last-frame or 4K from the August developer post as if they were this button.
4. Set duration from the signed line, not from a vibe. Google says you can choose the exact duration so the clip aligns with story and voiceover. If the VO is not live in Vids yet, time the picture to a VO you will lay in later.
5. Extend only after the first beat is on-model. Extension is for keeping lighting and character. It is not a second location.
6. Do not wait for the TTS sentence. Gemini 3.8 Flash-Lite text-to-speech is coming soon. Record or generate VO on a live desk, or ship silent with burned-in type.
7. Keep SynthID in the file you archive. It is how Google says viewers can verify AI origin. It does not replace your claim review.
8. Compare motion models on the same still if the first take drifts. Use Compare rather than hopping vendors because one export looked sharper.
9. If you already have a real take, restyle it in the NLE (Runway’s Resolve plugin is the 23 September example). Do not regenerate the hero from a prompt to save a round trip.

Worked example: a restock promo, not a “free HD” dump

Say the job is a restock announcement for a two-pack that already has an approved bottle still.

Do not: open vids.new, prompt “cinematic 1080p ad for our vitamin C serum, make it look expensive, add a confident voiceover,” extend twice, and upload the first HD file.

Do:

1. Lock one claim and one ask on the AI script generator. Example shape, not a magic prompt: Same bottle. Same seal. Restock Friday. CTA: Tap the two-pack.
2. Split that into one clip-length beat. If you need camera language, steal a shot type from prompt ideas — do not paste the whole brand story.
3. Generate motion from the approved still on image to video. If you use Vids, treat Create AI videos as the motion pass, not the copy pass. Set duration to the spoken line you actually have, not to a coming-soon TTS.
4. If the still warps, stop extending. Change one variable — camera or lighting — and retry. If you need a different motion model on the same still, Compare.
5. Lay VO you can defend. Google’s 100+ language TTS is not the live path in the 23 September post. Private YouTube export from Vids is fine for review. It is not a media buy.

That is the whole product. Omni 1.1 made 1080p cheap in a Workspace tab. It did not make the restock true.

FAQ

Is Gemini Omni 1.1 in Google Vids free? Google’s 23 September post says anyone with a Google or Google Workspace account can generate high-quality videos at no cost using Omni 1.1 Flash. The same post points personal users to Google AI plans for more access, and Workspace Business/Enterprise for expanded pools. A free floor is not an unlimited production budget. Google did not publish a generation count there.

Does this replace a video script generator? No. The launch is generate, duration, extension, 1080p, and SynthID. In-Vids text-to-speech is coming soon. Lock the line on the AI script generator first. Kapwing already splits Script Generator from Video Generator in its nav for the same reason.

Can I quote a max clip length or 4K from this Vids launch? Not from the 23 September Vids post. It confirms 1080p generate and upscale of existing AI clips, plus duration control and scene extension, without a published max length. 4K and first/last-frame language live on the 27 August Omni 1.1 Flash developer post for Flow/API. Keep the desks apart.

What is SynthID doing on these clips? Google says every Omni 1.1 clip in Google Vids includes an imperceptible SynthID watermark in the frames so viewers can verify AI origin. Treat it as provenance. It does not approve your claim, and it does not mean the clip is safe to run as a paid ad without review.

Should I use Google Vids or a dedicated AI video generator? Use Vids when the job is a Workspace-browser generate with HD scenes and a timeline you already live in. Use a dedicated AI video generator or image to video when you need a named motion model, a locked still, or a compare pass. Use VEED only if you actually want their script-plus-narration one-shot — and you still have to accept the script. If the footage already exists on a Resolve timeline, restyle it; do not regenerate it because HD got cheaper.

Is the voiceover live today? Not per Google’s 23 September wording. Gemini 3.8 Flash-Lite TTS in Google Vids is coming soon, 100+ languages. Until that ships, narration is a separate desk.

Free 1080p changed the cost of a draft. It did not change the order of operations: script, still, then generate.