Как бесплатно генерировать видео с помощью Gemini Omni 1.1 топовый ИИ для генерации роликов — Transcript
Full transcript
- 0:00Hi. My name is Evgeny Kovalenko, and on
- 0:03September 23rd, so yesterday, Google
- 0:05announced that anyone can now create
- 0:07stunning videos. These are, how can I
- 0:11say, amazing, yes, such delightful,
- 0:16stunning videos in HD format, that is,
- 0:191.280 x 720 with Gemini Omni. Yes,
- 0:24that's their model, it's the top model
- 0:27for video generation. They integrated
- 0:30their latest model into Google, so
- 0:33anyone with a Google account can create
- 0:36videos for free. It's free. That is,
- 0:40anyone with a Google account can do
- 0:42this for free since yesterday. I'll
- 0:45quickly run through the release and try
- 0:47to see what they updated there, how to
- 0:50use it. Yes, Google writes in its
- 0:52release that they are sharing the magic
- 0:54of video creation with anyone. Yes, it
- 0:57is expanding access to Google. Now
- 1:00anyone with a Google account or Google
- 1:02Workspace account can generate
- 1:04high-quality videos. No cost using our
- 1:08latest model, meaning no financial
- 1:11outlay, using their latest Gemini Omni
- 1:131.1 Flash model. I went to Artificial
- 1:17Analysis Arena—they took my word for
- 1:20it, right? So, we can see the
- 1:22leaderboard. Gemini Omni Flash is a
- 1:26slight lead over Omni 3.0. It's not
- 1:29specified which version, but we can see
- 1:32it's from Google, meaning it's either
- 1:3411 or 1.0. So, we can see Google is in
- 1:37the lead. Returning to the release, the
- 1:40model and its creative settings. Yes,
- 1:43it doesn't matter whether you're
- 1:45launching a business, uniting a
- 1:46community, or presenting your product.
- 1:51You can use Google's latest advanced EI
- 1:54model to create inspiring videos
- 1:57without professional editing skills or
- 2:00a big budget. Here, in the background,
- 2:03we can see all sorts of promotional
- 2:05opportunities from the release. Yes,
- 2:08look how people can generate and edit
- 2:10video scenes using OMNI 1.1 in GoogleS
- 2:13at no cost. They seem to point out in
- 2:17every block that it's free, no money,
- 2:20for everyone. To use the generation
- 2:23feature, go to the wiids.new website,
- 2:25and everything will happen there. I'll
- 2:28go there now and try it out. And here,
- 2:30going down a bit, yes, they have a good
- 2:33look at Omni 1.1 in GoogleS. In general
- 2:36, this model—I filmed a separate
- 2:37video, but I'll quickly say it here.
- 2:40The first feature they highlight is
- 2:42scene expansion with, uh, smooth, yes,
- 2:44transitions. That is, you can make a
- 2:48scene longer, extend it, while
- 2:50maintaining the visual context,
- 2:51maintaining the lighting, maintaining
- 2:53the characters, maintaining the way
- 2:55everything looks, yes, and, uh, the
- 2:57consistency of the environment. The
- 3:00second thing they highlight is the
- 3:02ability to control how long the scene
- 3:05should be, right? That is, choose the
- 3:08duration of your generated clip that
- 3:11suits you, yes, so that it clearly
- 3:15matches your story and your voiceover.
- 3:19The third thing they highlight is the
- 3:21ability to create new videos and videos
- 3:23of generated scenes in Full HD. It's
- 3:27strange, they say HD here, but Full HD
- 3:29here. Yeah, right. So 1920 x 1080 is
- 3:31Full HD. HD is just 1.280 x 720. Well,
- 3:37they indicate that it's possible. It
- 3:39doesn't say it's free, but it does say
- 3:41it's possible. And upscaling, yes,
- 3:44meaning you can already improve, expand
- 3:47, and enlarge existing AI-generated
- 3:49videos, yes, using their features, yes,
- 3:52video enhancement, expansion. Then they
- 3:55say it's created for all audiences. And
- 3:59then they say that every video
- 4:01generated with Gemini Omni has a Synth
- 4:03ID embedded. Synth ID is an invisible
- 4:06watermark, yes, which is embedded in
- 4:09every frame of the video. It's being
- 4:12done so that if someone, anyone, wants
- 4:15to check whether a video was generated
- 4:18using artificial intelligence, they can
- 4:22do so thanks to this built-in watermark
- 4:25. Yes, this is simply designed to
- 4:28create and share transparency, so that
- 4:31everything is, well, so that people
- 4:33know what's real and what's generated,
- 4:35right? Let's go to the very bottom of
- 4:39the release, it says again, go to
- 4:40visit.ne new, I click on "vits new."
- 4:43I'm redirected to a website where
- 4:44everything is quite simple. There's a
- 4:46window with choices, yes, I don't think
- 4:47I'm blocking anything. The only thing
- 4:49here is simply creating a video, right?
- 4:53Among the new features, GoogleS has
- 4:55created AI Videos, which is video
- 4:58creation using artificial intelligence,
- 5:00avatar creation, and document
- 5:02conversion. So, if you have a Word
- 5:05document, you can upload it and turn it
- 5:07into a video. There are also video
- 5:10slides. There's just a standard video
- 5:13recording. There's the ability to
- 5:15upload, uh, videos, just store a
- 5:17template, yes, that is, use, uh,
- 5:20standard templates for beautiful videos
- 5:22with transitions, like here. And
- 5:25there's Edit Videos. It's a pretty cool
- 5:28thing, as you can see. The girl snaps
- 5:30her fingers, and her style changes
- 5:32every time. Yeah, I think you've seen a
- 5:35similar video. It's only for a premium
- 5:37account. Basically, they didn't promise
- 5:41that video editing would be free or for
- 5:44anyone and so on. But let's move on to
- 5:47creating a video. I'll click Create
- 5:49Video. This is the interface that opens
- 5:51up for me, yes, where I can use it. The
- 5:54studio is like this, yes, where I can
- 5:55use the tools. So, I have Create,
- 5:57describe your video. You can add
- 6:00ingredients like brand images,
- 6:02characters, and more. Yes, you can
- 6:05create an avatar right away. And the
- 6:07ingredients are up to three images,
- 6:09which are described here, right? So,
- 6:12these could be parts of clothing, it
- 6:14could be some kind of appearance, and
- 6:16so on. We switch at the bottom. We can
- 6:20choose landscape—this is a landscape,
- 6:23uh, photo, like this, yeah, wide—and
- 6:25portrait—this is a vertical
- 6:27resolution. So, as I said, you see,
- 6:31they didn't mention in the release that
- 6:331080 is free to generate, but they did
- 6:36mention that it's possible. So, this is
- 6:391.280 by 720. The video model is the
- 6:43latest. Omni has no other choice here.
- 6:46And you can choose duration, which is a
- 6:48duration of no less than 3 seconds and
- 6:50no more than 10 seconds. We can
- 6:54basically leave these settings and try
- 6:56something and try generating something.
- 7:00Here I have an empty frame for 5
- 7:02seconds. You can click the plus sign,
- 7:07so you can insert a generated video
- 7:09into this empty frame, you can delete
- 7:11it, you can put some real video, and
- 7:14then build it up with a generated one,
- 7:16right? The plus sign adds another frame
- 7:20for 5 seconds. And here in the middle,
- 7:23we can create a transition, that is,
- 7:25add a transition. And you can delete it
- 7:28: just select it, press delete, and
- 7:30it’s deleted. I select it, press
- 7:32Delete, and everything is deleted.
- 7:34There’s an empty space, I select it,
- 7:36yes, I click on AI video. And now I can
- 7:38start generating. That is, if you
- 7:40don’t have this, then you need to
- 7:42click on the plus sign, create, select
- 7:44it from the left sidebar, in the right
- 7:46sidebar, click on A video, and start
- 7:48generating the video. There is also the
- 7:51option to simply create a video from an
- 7:53image, yes, Animate. That is, we click
- 7:56on Animate, and we need to add an image
- 7:58here and describe how we want this
- 8:01image to be animated. Let me click on
- 8:03add image. I’ll upload this image
- 8:07here. It’s Henry Cavell. Cavil. Cavil
- 8:11Superman there. Who else is he? The
- 8:13Witcher. Anyway, Kenny, who knows,
- 8:16knows. Who doesn’t know, well, what
- 8:18can I do? I generated it, by the way, I
- 8:19generated it on the RF model. A fairly
- 8:22old model f, but it handled it quite
- 8:25well. And now I'll set it, that is, we
- 8:28have 10 seconds left. I hope, well,
- 8:31let's say 10 seconds, I guess. I think
- 8:33it will be faster, but it's better to
- 8:35leave it just in case. 10 seconds. I
- 8:37wrote a prompt. And I wrote a prompt
- 8:39not just so that it would do some
- 8:41animation, but also so that it would
- 8:43speak, yes, that is, so that it would
- 8:45not just do something in the frame, but
- 8:47also speak and write text in Russian.
- 8:50Let's see how this works in Google. I
- 8:53send it, and your video starts
- 8:56generating. 100%written. Ah, so, I
- 9:00can't do this with him because it
- 9:04determined his appearance. and told me
- 9:07that I was violating the terms of use.
- 9:09Okay, fine. Let me try to take a simply
- 9:12generated woman who doesn't really
- 9:14exist. So, I said I'd upload a girl.
- 9:17And let's try animating her. Here I
- 9:21have her completely, damn, it's a
- 9:23square landscape. I wonder if it will
- 9:26turn it from square to rectangular or
- 9:29not? That's the question. The girl is
- 9:32completely generated, so there
- 9:33shouldn't be any problems with
- 9:35generation. I used the "create a video
- 9:38from a prompt" function, but that was
- 9:40also a bit of a mess. I'll only share
- 9:43this result for sponsors because I
- 9:45can't put such things in a regular
- 9:47video or on the cover. YouTube will
- 9:49simply censor me and won't show it to
- 9:52anyone. That means it will have 10
- 9:53views. There you go. Um, so if anyone
- 9:55is interested in the result, please
- 9:57sponsor. There are many other videos
- 9:59out there, I'll keep trying. Look,
- 10:01decide for yourself. No one is forcing
- 10:04you, or go to Google and try, test. I
- 10:07entered the prompt and sent the video
- 10:09for generation. It should now generate
- 10:11without any problems. In principle, the
- 10:13video is generated pretty quickly. So,
- 10:15while I'm talking, you can see the
- 10:16percentages flashing on the screen. As
- 10:18soon as it reaches 100, the video will
- 10:20be ready. I created a video generator,
- 10:23a specific action for the girl to
- 10:25perform, and for her to say a phrase in
- 10:27Russian. Now that's quite an
- 10:29interesting point. If I can do it. If
- 10:32not, well, then for now, it will
- 10:34probably be in English as standard. But
- 10:36in general, Gemini, Google, they
- 10:39generally brag and boast about having
- 10:42the ability to generate videos in
- 10:44seventy languages, 110 languages. So,
- 10:47if that's true, you can, uh, give it a
- 10:50try. Let's listen.
- 10:52I'm a fan of Evgeniy Kovalenko.
- 10:54Subscribe to his channel so you don't
- 10:56miss new videos about.
- 10:57Yes, only such fangirls can be
- 10:59generated for me using AI.
- 11:01Hi,
- 11:02I'm a fan of Evgeniy Kovalenko.
- 11:04Subscribe to his channel so you don't
- 11:06miss new videos. It was generated with
- 11:08the help of some amazing quality,
- 11:10meaning everything that was on the
- 11:12video is preserved. She gestures, she
- 11:14speaks beautifully, the voice is
- 11:16perfectly placed. Basically, as you saw
- 11:18before, I don't think you need to
- 11:19listen to it 50 times. Basically, you
- 11:21can rewind and listen to it. There's a
- 11:24Jiminy marker, and we can insert this
- 11:26frame down here. It's ten seconds long,
- 11:29as I said, it will take less than a
- 11:32second, but overall, honestly, I'm
- 11:34happy with the result of the generation
- 11:37, I'm happy with the quality, I'm happy
- 11:40with the way she speaks, the way she
- 11:43gestures, how beautifully it's all
- 11:45created, how it sounds. Well, honestly,
- 11:48my impression is, well, probably a nine
- 11:52out of ten. Well, close to perfect, yes
- 11:55. There is this one moment here that I
- 11:57didn't like. Now, just a second. That
- 12:00is, yes, this initial moment, when it
- 12:02somehow somehow cuts off. Well,
- 12:06basically, there's a certain trick to
- 12:08it, yeah, it's like a setup, like, with
- 12:11a frame behind the scenes, yeah, like a
- 12:14frame before, something that shouldn't
- 12:17have been there, but it's kind of
- 12:19inserted specifically here, yeah, to
- 12:22create this kind of style, yeah, but
- 12:24overall, it's pretty cool and
- 12:26well-written, so I don't know, I really
- 12:29like it. I generated it like this. The
- 12:33first video, it blocked me because of
- 12:34Henry Cavel's appearance. The second
- 12:37video, it generated for me, I'll only
- 12:39show it to sponsors. And the third, uh,
- 12:42you see it generated right before your
- 12:44eyes. So far, it doesn't ask for any...
- 12:47That is, I created the video through
- 12:49create, generated it completely and
- 12:51created it through animate, uploaded it
- 12:53. Yeah. Uh, so far it looks pretty cool
- 12:56. You can use this girl as an avatar.
- 12:59See, it stays right here. It's labeled
- 13:01A image, so we can link to it with a
- 13:03dog. So, any asset, any of yours, how
- 13:06should I say, assets, yes, you can
- 13:08specify them using the "at" sign, yes,
- 13:11so instead of constantly uploading the
- 13:13same photo, the same bag, the same
- 13:15clothes, the same person, the same
- 13:18talking head, and so on, you can press
- 13:20the "at" sign and write the caption. So
- 13:23, we have one image here, but we can,
- 13:24in principle, rename it or, in
- 13:26principle, access it as one image,
- 13:27right? So, I'll type like this with the
- 13:29"at" sign, and it will prompt me. And
- 13:32in principle, they won't all line up
- 13:34and prompt you one by one, right? Again
- 13:37, we can, for example, extend this
- 13:39video, yes, by 15 seconds. We can
- 13:42extend it, we can add another video
- 13:44here, create a transition. So, we can
- 13:46generate additional videos and create a
- 13:48transition between them, a smooth
- 13:50transition. Yes, something like, I
- 13:53don't know, it splits the screen or it
- 13:55zooms in and out—depending on how,
- 13:56what you choose. Yes, when I press "
- 13:59dissolve," one transitions into another
- 14:01. Fate is fading, yes, it even
- 14:03basically suggests on the screen what
- 14:04transitions there can be, right? That
- 14:06is, "zzolv," well, you can't see it
- 14:08because one transitions into another,
- 14:10and "fay" is from light to dark. "Push"
- 14:12pushes the slide. Well, basically, you
- 14:14can experiment. It's a pretty simple,
- 14:17easy, and intuitive studio. Volume
- 14:21control, play the video at the top,
- 14:22insert it at the bottom using the "
- 14:24insert" button. Yes, we press "insert,"
- 14:26it inserts. "Extend" expands the clip,
- 14:28right? So, seamlessly, so-called,
- 14:31expands it, continues it, right? That
- 14:34is, we press "extend" and describe how
- 14:36we want the video to continue. "
- 14:38Recreate" simply regenerates it, and "
- 14:41remove" deletes it. And, of course, the
- 14:43most important thing is how do we save
- 14:46it? Let's click on the file. And we can
- 14:52download it. We can download this video
- 14:54in MP4 format. And it's preparing to
- 14:57download. I would really like to
- 14:59download it and use it in my video so
- 15:02that people can watch. Yes, the video
- 15:04has downloaded, here it is, 3.5 MB, 10
- 15:06seconds. If you liked this video, you
- 15:10can give it a thumbs up, subscribe to
- 15:12the channel and click the bell so you
- 15:14don't miss new videos on this channel.
- 15:17And that's all. Thanks for watching.
- 15:19See you later. Good luck. Bye.
- 15:21Hello. I'm a fan of Evgeny Kovalenko.
- 15:24Subscribe to his channel so you don't
- 15:26miss new videos about And good luck.
About this transcript
This page contains the full transcript of Как бесплатно генерировать видео с помощью Gemini Omni 1.1 топовый ИИ для генерации роликов by Yevgeniy Kovalenko, generated from the public captions YouTube serves with the video. The transcript has 2,466 words across 367 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.
What you can do with it
Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.
Free YouTube transcript tool
YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.