YT YouTube2Text

YouTube2Text → Transcripts → The ONLY Free & Unlimited AI Video Generator Left

The ONLY Free & Unlimited AI Video Generator Left — Transcript

by Malva AI · 3,045 words · 418 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:01I've found the ultimate free AI. With
0:03it, we can make unlimited 10 second
0:05[music] videos with no watermark and
0:07audio included, and it works on both
0:09computers and phones. It also has an
0:11agent mode, the kind the best models
0:13like Claude or ChatGPT have. We can
0:16leave it working while we do other
0:17things and it takes care of all the
0:19tasks. For example, with a very simple
0:21description, I asked it to make a film
0:23about this samurai, and all I told it
0:26was what I wanted him to do in each
0:27scene. [music] In agent mode, it first
0:29defines what the main character and the
0:31setting look like. Then it generates all
0:33the videos in one go without us having
0:36to make them one by one. It also adds
0:38the references to each video, so the
0:40character and the setting are always the
0:42same and don't change from [music] shot
0:43to shot. It gave me all the videos at
0:45once, and look how well the result came
0:47out.
0:50>> My time has come.
0:59>> So you can use this to make things that
1:01run for several minutes with a lot of
1:02videos at once, like those long videos
1:05that have lately been getting millions
1:06of views. Or with a method I'll show you
1:08in a bit, you can use its reasoning
1:10model to create more complex things,
1:12like ads you can sell to brands. Maybe
1:15you've already tried this AI and it
1:16wouldn't let you in because it causes
1:18quite a few problems when you try to use
1:20it. Today I'll show you how to solve all
1:22of them. It doesn't matter if it blocks
1:24you with an error, won't let you sign
1:25in, or asks for a credit card. There's a
1:28legal and safe way in without giving
1:30your card details. We're going to fix
1:32all of this so that wherever you are and
1:34whatever you want to create, you can use
1:36this AI completely free. Let's open Muse
1:39AI. In a second, [music] I'll give you
1:41all the direct links and the prompts
1:43from today's video, so you only have to
1:45copy [music] and paste them. Once we're
1:46in the interface, the first thing we do
1:48is sign in. If you already get an error
1:51at this point, I'll show you how to fix
1:52it later along with the other common
1:54errors. You sign in with a phone number
1:56or an email like on any other platform.
1:59Then it asks to fill in these details.
2:01[music] And here comes one of the first
2:03problems. To confirm you can use the
2:05platform, it asks for a credit card.
2:07Luckily there's another way to sign in
2:09right inside this same platform [music]
2:11and we won't have to use anything
2:13external. At the top right there's a
2:15gear icon, so click it. Under general
2:18click meta account, which is the only
2:20option you'll get. That takes us to
2:22another section showing the email or
2:23phone number we signed in with. [music]
2:25Click it and in the next menu click it
2:27once more. In this case the email. At
2:30the very bottom of this one click add or
2:32[music] edit contact info. In the next
2:34menu go to add new contact info and then
2:37[music] add mobile number. And this is
2:39where we add our phone number. You can
2:40click change to pick your country, type
2:43in your number and click next. [music]
2:45It asks for a confirmation code that
2:46gets sent to your phone. Paste it in and
2:48wait until it says mobile number added
2:50at the bottom. Now you can close this
2:53and go back to the tab we were on
2:54before. Press F5 to reload the page and
2:57we're in without the error and without
2:59entering a credit card using only the
3:01platform's own settings and no outside
3:03tools. Now comes the best part because
3:06in here we can do everything from text
3:08and [music] images to videos and its
3:10reasoning model. You won't need to pay
3:11for any other AI at any point as long as
3:14you learn to use this one properly. In
3:16the first chat it asks what it should
3:18call us. You can put any name here and
3:20I'm going to write Malva. Then it asks
3:22what we want to call it. This matters
3:24because like I told you earlier, what we
3:26have here is practically an assistant.
3:29If you look at the top right there's
3:30Muse which will be working on everything
3:32we ask for. I'll also show you later how
3:34to tell when it has understood you
3:36properly or when it's giving you an
3:37error. Next it asks if we want to
3:40connect apps like Gmail or the calendar.
3:42We can connect it to other tools [music]
3:44so it sends emails for us, tells us if
3:46someone has sent us something important
3:47lately or reminds us when we have
3:49[music] an appointment coming up. This
3:50is very useful especially on your phone.
3:53Like I said at the start, if we go to
3:55this option here and click download apps
3:57we can get it on both iOS and Android
3:59and have our assistant on the phone. But
4:01let's get to the videos. There are a lot
4:03of ways to generate them here, but for
4:05me, only one is the right one. The first
4:08is the one we all know. We can do it in
4:10the main chat, or we can add a new one
4:12under side chat to keep our work
4:13organized. Here, we do the usual thing.
4:16We add some text describing the video we
4:18want and the aspect ratio, and we click
4:20generate. The result is very good.
4:33A 10-second video in good quality, and
4:35the audio sounds right and matches
4:37what's happening. But we can take it a
4:39step further because certain commands
4:40directly improve our videos. To start
4:43with, we can add the line lipsync and
4:45then, in quotes, [music]
4:46exactly what we want the character to
4:48say. Generate it now and look at the
4:50result.
4:53>> I have to find him before it's too late.
5:01>> That way, we can build a story where our
5:02character talks and says exactly what we
5:04want. And we don't even have to do this
5:06ourselves because with the agent mode
5:08we'll see later, it takes care of
5:10creating all the dialogue and all the
5:12videos in one go. Also, don't worry
5:14about how much you're generating. If you
5:16go to settings, you'll see it says
5:18usage. We've generated a lot of things,
5:20and it's still fairly low. But this
5:22isn't even the limit on videos. It's
5:24just there so we can keep track of how
5:26much we've used [music] the tool. I've
5:27reached 100% here many times and kept
5:30creating a huge number of videos. Even
5:32so, you can see it resets every few
5:34days. So if you get an error, you just
5:37wait a little and keep going. [music] In
5:38a minute, I'll show you how to make
5:40better free videos than the ones we've
5:42just seen. What if we want to make viral
5:44content to upload to places like YouTube
5:46Shorts or Reels? You've been asking me
5:48for this a lot in the comments. How to
5:50make those Shorts with millions and
5:52millions of views using AI? Well, I've
5:54got two [music] options for you. And the
5:56first is with Higgsfield, today's
5:58sponsor, which has just released a new
6:00feature called Higgsfield Shorts.
6:02[music] It lets you start from a lot of
6:03videos that have already gone viral and
6:05change them so the star is you, a
6:07character or your own products. [music]
6:09As you scroll down the feed, you'll see
6:11these presets along with how much
6:13attention each one got. To make your
6:15own, there's a guide that explains how
6:16it works [music] and you have to connect
6:18Higgsfield to Claude. It's very simple.
6:20Just this button and a short tutorial.
6:23Continue, add, connect, [music]
6:25and allow. And then you've got a very
6:27strong reasoning model helping you make
6:28your videos. Now, pick a short that
6:31stands out to you from here or one
6:32that's come up in your own feed and
6:34click recreate in Claude. You can see
6:36the code gets added by itself [music]
6:38and Claude already knows exactly which
6:40video you mean. Then add photos of your
6:43characters or your products and ask
6:44[music] it to recreate the video with
6:46them. You wait a little, keep refining
6:48what you need by talking to Claude, and
6:50your video's ready.
6:52>> Doubt [music] shows up every morning.
6:56So do you.
6:58Show up on the days [music] you don't
7:00feel ready,
7:02the days nobody's watching,
7:05and give it everything you've got.
7:06[music]
7:08>> And I recommend you don't stop at one
7:10final result. Give the video your own
7:12style [music] and ask Claude to change
7:14the text, the shots, and the sounds
7:16until it's perfect. And the links in the
7:18description. Thanks to Higgsfield for
7:20supporting this video. The second option
7:22is right inside Muse and it uses agent
7:24mode. For me, agent mode is the best way
7:27to work here instead of having to add
7:29every detail ourselves. So let's go
7:31through the samurai film from earlier in
7:33depth. [music] We don't add each clip of
7:35the video prompt by prompt because that
7:37way we risk every one of them coming out
7:39with different content. We tell it what
7:41we [music] want, a scene of six clips of
7:4310 seconds each in 16 by 9, so it's
7:46horizontal with the same character in
7:48every clip. We describe exactly what our
7:51main character looks like. Then with a
7:53very simple prompt, without having to
7:55write everything out or add a photo for
7:57each clip, we say how we want each one
7:59to be. In some of them, we use
8:01techniques we saw earlier like lip
8:02[music] sync. And at the end, we give it
8:04a message with a few things that have to
8:06be followed no matter what. This is also
8:09where the second option for shorts comes
8:10in. If you want, when you add this
8:12prompt, you can tell it you want a viral
8:14video in vertical format for YouTube
8:16shorts. That way, it analyzes the best
8:19retention techniques so people stay to
8:21watch your video. It creates a vertical
8:23video and that takes care of your
8:25shorts. I'll leave both this prompt and
8:27a template for making your own along
8:29with the direct link to Muse AI for
8:31computer and phone on my website in the
8:33description. Just go to PDF guides,
8:35click the one with the thumbnail of the
8:37video you're watching, and download it
8:39all for free so you only have to copy
8:41and paste. When this emoji shows up
8:43under our message, it means Muse has
8:45understood the task perfectly and is
8:47already working on it. If it doesn't
8:49show up, it's important to check
8:50carefully because there may be an error
8:53or it may not have understood your
8:54instructions. You can see it's already
8:56working and it goes through different
8:58stages that we can check in the menu on
8:59the right. [music] Here, we can see what
9:01it's doing at each moment and it has
9:03different modes. This mode with the orb
9:06seems to be more creative [music] and
9:07this one with a computer seems more
9:09practical. That's because it has
9:11different agents that specialize in
9:13different activities and it uses all of
9:15them at once to reach the best possible
9:17result. And since [music] everything
9:18it's doing is broken down here, we can
9:20see if it got a part wrong, which we
9:22normally wouldn't notice with another
9:24AI. We can correct it right here so the
9:26process goes better. The more we ask
9:28for, the longer it takes. And I
9:30recommend generating them six at a time
9:32at most to keep mistakes under control.
9:34Take the samurai clips. [music] You
9:36already saw them earlier, so I won't
9:37play them again. Some of them came out
9:39with mistakes [music] and while they're
9:41not that bad, we can't leave them in if
9:43we want to make a serious product.
9:44[music] Like I said before, this tool is
9:47unlimited, so we can tell it we didn't
9:49like clip number three because it has a
9:50mistake and that we want it regenerated
9:52and fixed. It understands and it
9:54regenerates only that. But how do we
9:57continue this story? We already have six
9:59clips, so how do we make the next part
10:01without the character changing? [music]
10:03It's very easy because this is an agent
10:05mode and it keeps everything it creates
10:07in its context. [music] That lets us do
10:09a few different things. We can it for an
10:11image of the samurai from the video, add
10:13that back as a reference and tell it to
10:15use the same one in every video. Or we
10:18can do something else that goes with the
10:19next thing I wanted [music] to show you,
10:21video editing. Look, let's take this
10:23clip as a reference and ask it to edit a
10:25few things. First, let's ask it to edit
10:27only the setting and keep the main
10:29character the same. We send the prompt
10:31with the video uploaded. This one takes
10:33much less time because it's a single
10:35video and a few seconds later we have
10:37our samurai, but this time galloping
10:39through the snow. Or we can do the
10:41opposite and ask it to keep the setting
10:43exactly the same but change the main
10:45character. We send it and now our main
10:48character is a Spartan in the same
10:49meadow as before. You can use this to
10:51fix scenes you're not quite happy with
10:53or to make your own version of a video
10:55that catches your attention. Or you can
10:57ask for a continuation of a video with
10:59the same main characters. We can upload
11:01a video so it understands the setting,
11:03some images of what our character looks
11:05like, and all the content we need, and
11:07it handles the rest. Now, for those ads
11:10you can sell to brands and making them
11:11is very simple. We open a new chat or
11:14the main chat and you can see they all
11:16get saved here. Then we ask for exactly
11:18what we want to do. We want to create
11:20[music] an ad for this product and we
11:22want it to go super viral. The idea is
11:24that it reasons through the best
11:26techniques for making people like an ad.
11:28Once it has them, it writes several
11:29prompts and turns them into video to
11:31create the ad for the product we're
11:33sending. We add a reference for what we
11:35want the ad to [music] be and it starts
11:36reasoning and creating all the content
11:38it needs. I'll also leave this prompt on
11:40my website in the description. A few
11:42minutes later it has finished all the
11:44planning and the prompts. If you want as
11:46the results come in, you can ask it to
11:48change [music] any detail and your ads
11:50are ready.
12:09And we don't even have to bother
12:10publishing them on each platform.
12:12[music] We can go to settings at the
12:13bottom left and then to connectors in
12:15this menu where there's a huge number of
12:17apps for productivity, social media,
12:19[music] health, and business. It can
12:21then handle things like publishing our
12:23videos on these social networks or if we
12:25make a logo for a brand with the image
12:27generator, which we haven't looked at
12:29but is also pretty good, it can upload
12:31it to Canva, for example. [music] If
12:33we've created some code, it can upload
12:34that to GitHub all without paying
12:36anything. [music] You can see there's an
12:38upgrade option here, but I get the
12:40impression this is still in beta
12:41because, like I said, I've hit 100% many
12:44times and kept using it. [music] So beta
12:47or not, I recommend you use it. In the
12:49menu on the left, we also have sections
12:51like ideas. If this video didn't make
12:53clear everything it can do, you can look
12:55through those ideas here and they're
12:57based on the kind of things we've done,
12:59not picked at random. For example, if I
13:01use this tool to make my videos or my
13:03scripts, it would show me things here
13:05for improving [music] those scripts. So
13:07as we keep using it, we'll even get
13:09better at everything without [music]
13:10realizing it. If we scroll down, there
13:12are also ideas for other very
13:14interesting things, even for improving
13:16parts of our daily life, like our
13:18finances, what we buy, and how we relate
13:20to people. There's also a feed where we
13:22can ask it for news on certain topics or
13:25just scroll down and see what it has
13:26decided is relevant enough to show
13:28[music] us. Of the common errors you
13:30might get, I think the most important
13:31one is at sign in. It can give an error
13:34depending on your region, although it
13:36seems that's no longer happening. If it
13:38does happen to you, [music] you just
13:39need to use a VPN and any free one
13:41works. I'll leave the one I use in the
13:44description with a referral link, but
13:45you can use any other free one. Another
13:48common error I ran into while generating
13:50videos is that I'd ask for something
13:51totally normal and [music] it would tell
13:53me it couldn't generate it. That wasn't
13:55because of any video limit. It said it
13:57couldn't generate that content. [music]
13:59To fix that, I recommend opening a new
14:01chat and giving it this prompt I've
14:03already prepared. That prompt asks it to
14:05turn every prompt we send in that
14:06[music] chat into one that follows its
14:08policies. That way, a normal request
14:10like that goes through and we can
14:12generate the videos we need. For any
14:14other problem, if it's something
14:15specific, I recommend asking the chat
14:17itself how to fix it. If it's more
14:19complicated and the chat [music] can't
14:21help you, leave it in the comments and
14:23I'll try to help you as soon as I can.
14:24All the links and prompts from today are
14:26on my website in the description. See
14:28you in the next one and thanks for
14:29watching.

About this transcript

This page contains the full transcript of The ONLY Free & Unlimited AI Video Generator Left by Malva AI, generated from the public captions YouTube serves with the video. The transcript has 3,045 words across 418 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 choose a caption track available on the video.

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, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.