Stop Wasting Credits on AI Video Generators Like Seedance 2.5 — Transcript
Full transcript
- 0:12No, no, no, no.
- 0:15I need credits. I need to find credits.
- 0:30So, there are three methods of how you
- 0:31can save yourself money on AI video
- 0:33credits that we'll be covering in this
- 0:35video. We'll be using Higgs with for
- 0:37this as in my opinion, this is one of
- 0:39the best AI video generator tools out
- 0:41there and they have created a new method
- 0:43that will save you a lot of credits
- 0:45which I'll be covering later. I will
- 0:47leave a link to them in the description
- 0:49down below. The first method is choosing
- 0:51the right model and this often gets
- 0:53overlooked by a lot of people. Everyone
- 0:56nowadays thinks they have to use
- 0:57Seedance 2.5, but to be honest, it's
- 1:00expensive as [ __ ] A different model can
- 1:02easily achieve 80 or 90% of the same
- 1:05result and it's [music] a lot cheaper.
- 1:07So, let me show you an example of what I
- 1:09mean by that. So, I have one video that
- 1:11is generated using Seedance 2.5 and the
- 1:14other video is generated using an older
- 1:16model which is Kling 3.0 which is
- 1:19severely outdated by now. I won't tell
- 1:21you which one is which yet, but you tell
- 1:23me which one you like the most and which
- 1:25one you think is Seedance.
- 1:40Now, the first video was Kling 3.0 and
- 1:42the second video was Seedance 2.5. To
- 1:45me, these results are really similar,
- 1:48but this generation in Kling cost me 12
- 1:50credits while the same generation in
- 1:52Seedance 2.5 cost me 72 credits. That's
- 1:56the difference of 60 credits. So, if we
- 1:58compare the two, I can generate six
- 2:00videos using Kling 3.0 for the same
- 2:03price as one video on Seidance 2.5. And
- 2:06that's why you got to identify when you
- 2:08can use a cheaper model. Because some of
- 2:10the models might not be suitable for
- 2:12your generation. For example, let's take
- 2:14the topic of anime. I've done a lot of
- 2:17research on this. I've asked around and
- 2:19I've tried it myself. And in my
- 2:21experience, Minimax is the better
- 2:24solution to use here compared to
- 2:25something like Seidance. So, here I have
- 2:28another example.
- 2:42>> I am the AI lord.
- 2:59I am the AI lord.
- 3:00>> Here, the first generation was Seidance
- 3:022.5 and the second one was Minimax. I
- 3:05liked the result of Minimax a lot more
- 3:07and it's also a lot cheaper. The
- 3:10difference here in price is 180 credits
- 3:12for Seidance 2.5, while it has only 30
- 3:15credits for Minimax H3. And if you use
- 3:18the max version, it's not even that much
- 3:20more expensive. It's 37.5. And the other
- 3:23thing is, Minimax is pretty
- 3:25unrestricted, while Seidance 2.5 has the
- 3:28tendency to not allow a lot of different
- 3:30generations. Like, for example, if you
- 3:32want to make copyrighted stuff, you
- 3:34can't really do that with Seidance. But,
- 3:36with Minimax on the other hand, I made
- 3:38this.
- 3:40>> Rick, I I told you we can't use
- 3:42copyrighted visuals.
- 3:43>> Yes, we can, Morty.
- 3:44>> No, we can't.
- 3:45>> Yes, we can, you stupid human.
- 3:47>> We definitely can, guys.
- 3:49>> No, we can't.
- 3:54>> He's too young. Now, if you're just
- 3:56getting started with AI, it might be
- 3:57super overwhelming trying to know
- 3:59everything about which model you need to
- 4:01use. And that's why I made this website
- 4:04for you, where you can upload your own
- 4:07prompt to it, and it will tell you which
- 4:09model you should use. So, call this the
- 4:12AI video model selector. How it works is
- 4:15you simply add in your prompt. This is
- 4:17powered by Jeff, so it does the hard
- 4:19work for us. It's run by me, it's super
- 4:21cheap. I will just give this away for
- 4:22free. You put in your prompt, then you
- 4:25add in the reference that you have, and
- 4:27then you tell it the duration that the
- 4:28video will be. As you can see, if you're
- 4:30trying to make longer videos than, for
- 4:33example, 15 seconds, you only have a
- 4:35choice of these three models. If you go
- 4:37longer than that, you only have the
- 4:38choice to do something like SeaDance or
- 4:41One 3.0. So, let's do 10 seconds as this
- 4:43first simple test. What it will do is it
- 4:46will evaluate your prompt, and it will
- 4:48look through the database of the things
- 4:51that I've uploaded there, and then it
- 4:52will tell you which model will get you
- 4:54the best result. And then it's up to you
- 4:56to make the judgment call if it's worth
- 4:58it. For example, here it's only giving
- 5:0033% to do it with SeaDance 2.5. And it
- 5:04will also give you the expected credit
- 5:06cost. But for 35 credits cheaper, which
- 5:09is 45 credits in total, you can also use
- 5:11SeaDance 2.0. And [music] the budget
- 5:13pick would be One 3.0. So, it will give
- 5:15you those percentages, and if you are
- 5:17And depending on what budget you have,
- 5:19you can make your choice of that. And
- 5:21once you like what you see, you click on
- 5:22generate, and then it will bring you
- 5:24over to Higgsfield, where you can
- 5:25generate your video.
- 5:33Okay, now let's try a more complex scene
- 5:35to show you how this really works. So,
- 5:37now I added in a longer prompt with a
- 5:40lot more words, and it's kind of like a
- 5:42sitcom type of prompt. I also have a
- 5:45couple different references. And then
- 5:47I'm going to select the length of how
- 5:49long I want this video to be, which is
- 5:5115 seconds. For the resolution, I want
- 5:53to do 1080p. And now we're going to
- 5:55evaluate, and then it tells us that this
- 5:59might be best to do with Cidence 2.5.
- 6:01Again, it's the most expensive model,
- 6:03but if you want the absolute best
- 6:04result, you want to be using this. If
- 6:06you want to do 4K, it will automatically
- 6:09drop that model because 4K is
- 6:11unavailable, but you can get 1080p. So,
- 6:14based on this recommendation, let's
- 6:15actually generate this scene using
- 6:17Cidence 2.5. And now it has added in the
- 6:20references, it has added in the prompt,
- 6:23and then we can hit generate.
- 6:26>> I finally found a friend.
- 6:27>> Please, everyone knows you don't have
- 6:29friends.
- 6:30>> Oh, yeah? You sure?
- 6:32>> Yeah, I'm sure.
- 6:33>> So, look at this.
- 6:36Hey, I'm Bruiser, Dan's new friend.
- 6:40>> So, yeah. Looking at this result, having
- 6:42multiple different camera angles, three
- 6:44different characters talking. It's not
- 6:46too complicated in terms of camera
- 6:47movement, but I would say Cidence 2.5
- 6:50really nailed this. And that's how I'm
- 6:52hoping to save credits for you. But
- 6:54let's say we do something completely
- 6:56different. So, let me remove all of
- 6:57these references, and then we add in a
- 6:59different prompt. This time I'm going to
- 7:01do a motion prompt. This is going to be
- 7:03for a 2D kinetic typography and motion
- 7:07graphics, 50 seconds. So, I've added in
- 7:09my prompt. I have added in the
- 7:11resolution, all of that. We're going to
- 7:13evaluate it, and now it's telling us
- 7:15that Minimax H3 might be the best option
- 7:18to use here because it is super good at
- 7:20making these motion graphics. And if we
- 7:22generate this, we get a result like
- 7:24this.
- 7:28>> [music]
- 7:35[music]
- 7:40>> So, all we did, and by we I mean me and
- 7:42team, is fed the data that we had when
- 7:45using these models, and based on that,
- 7:47we built this tool for you, so you can
- 7:49hopefully save yourself some credits.
- 7:51Again, this tool is free to use. I will
- 7:53leave the link in the description down
- 7:54below. The second method to saving
- 7:56yourself credits is using draft mode.
- 7:59This is only available for students 2.5,
- 8:01and ever since this launched, I've saved
- 8:03a lot of credits since then. Because, as
- 8:05you can see, the 1080p version is quite
- 8:08expensive. But as soon as we switch over
- 8:10to the draft mode, we get the video for
- 8:1245 credits. The downside, it's only
- 8:15480p. You might wonder, like, why would
- 8:18you do that? So, for example, I used to
- 8:21generate a lot of 1080p videos, and I
- 8:24regenerated them until I got it right.
- 8:27Each generation cost me like over $10
- 8:30with students 2.5 1080p. Now, with draft
- 8:33mode, this cost has significantly gone
- 8:35down, and then once I found the perfect
- 8:38generation, all I had to do is just
- 8:40upscale it. Because draft mode is 45
- 8:43credits, and then it will cost you an
- 8:44extra 180 credits to upscale it. So, let
- 8:47me give you an example of what I mean by
- 8:49this. So, my goal is to kind of like
- 8:50create a Star Wars scene. I have me
- 8:53myself as a character sheet in there,
- 8:55and I have Bruiser Robot as a character
- 8:58sheet in there, and then I've added in
- 8:59my prompt. I'm not sure like if this
- 9:02will work, because it's quite a complex
- 9:04scene. What I want to see is I want to
- 9:05have a battle in space. I want to have
- 9:07some interaction in terms of talking
- 9:09with me and the other character, and I
- 9:11want everything to like flow nicely. So,
- 9:14instead of me just going straight into
- 9:16the 1080p mode, which will cost me a lot
- 9:18of credits, I first generated this
- 9:20draft. Now, that got me this first
- 9:23generation.
- 9:24>> [groaning]
- 9:27>> Wow, that was really close.
- 9:38You really need to learn how to speak.
- 9:41>> I've always known how to speak. It was
- 9:42just to piss you off.
- 9:43>> So, to me this doesn't really make sense
- 9:45because first we have the robot in the
- 9:47starfighter like this
- 9:49and then it's flying like outside of it.
- 9:52So, I won't be able to use this
- 9:54generation in my video. So, I have to
- 9:56regenerate it. But what I do now is I
- 9:58can just go back and I can just tweak
- 10:00this prompt. So, I look through my
- 10:02prompt. I find the parts that I want to
- 10:04tweak. I rewrite it and then I hit
- 10:06generate again again in draft mode. And
- 10:10that got me this generation.
- 10:15>> Wow.
- 10:16That was really close.
- 10:27You really need to learn how to speak.
- 10:29>> I've always known how to speak. It was
- 10:30just to piss you off.
- 10:32>> So, that is already a lot better. But
- 10:34now that I'm watching this, I am
- 10:35thinking like why don't I have the robot
- 10:38like sit outside of the spaceship
- 10:41instead of like in the cockpit with me.
- 10:43I go back into my prompt. I add in these
- 10:46changes and then I generate it again
- 10:49with the draft mode. And that generated
- 10:51me this video.
- 10:55>> Wow. That was really close.
- 11:07>> You really need to learn how to speak.
- 11:09>> I've always known how to speak. It was
- 11:11just to piss you off.
- 11:12>> And that's exactly how I want the video
- 11:14to be. Like this to me is great. Like I
- 11:16can use this. And then now that I have
- 11:19the final version, I had to pay one time
- 11:21a 240 credit fee to get it in the 1080p
- 11:25version. So, in total that would be 420
- 11:27credits. So, if you put them side by
- 11:29side, you would have saved yourself 300
- 11:32credits using the draft mode. But, if we
- 11:35compare this to how I used to do this
- 11:37the old way, is I used to have to
- 11:40generate this video three different
- 11:42times for 240 credits each, which will
- 11:45bring out the total price to 720 credits
- 11:48to now me generating this three times
- 11:50using the draft mode, which is 60
- 11:53credits per generation. The third method
- 11:55to saving yourself money is using the
- 11:58API. And in particularly, I'm talking
- 12:00about the Hixfeld API as they have
- 12:02launched this. Because if you're unsure
- 12:03if AI is something for you, or you find
- 12:05it expensive, and you don't have to lock
- 12:07yourself into a annual or a monthly
- 12:10membership, you can just pay for what
- 12:12you use. If you're unsure of what to
- 12:14use, then let me explain it to you.
- 12:16First, if you are using AI casually, and
- 12:20you're not spending all of the credits
- 12:21that you have in your monthly
- 12:23subscription, then get the API. You only
- 12:26pay for what you use. If every time you
- 12:28use AI and your or subscription is
- 12:31finished, you might be better off
- 12:33upgrading your subscription or seeing
- 12:35like what you can get with the API. But,
- 12:38if you are a heavy user, then I would
- 12:39still recommend using a subscription.
- 12:42I've actually made this graph for you,
- 12:43which exactly explains for which
- 12:45scenario it is best to either use the
- 12:47API or get a subscription and what
- 12:50subscription you should get. If you are
- 12:52making just one generation, API is
- 12:54obviously cheaper because you're wasting
- 12:57$49 on the plus plan and $310 if you're
- 13:00getting like a maxed out ultra plan if
- 13:02you don't use these credits. Now, if you
- 13:04make five generations, then the plus
- 13:06plan on an annual plan is cheaper. But,
- 13:09if you want a monthly plan, then it's
- 13:11about the same as the API. If you want
- 13:13to make more than five generations, so
- 13:15let's say 10 for example, you cannot
- 13:17even get that in the plus plan and we're
- 13:19talking about students 2.5 1080p
- 13:2215-second generation videos. So, these
- 13:24are long videos. Keep that in mind. You
- 13:26can also compare this to the Cidens 2.0
- 13:29and Kling 3.0. Over here, as you can
- 13:31see, most of the times the API is
- 13:33cheaper to use if you're using these
- 13:35models. But when you're using Cidens
- 13:362.5, which is the model that most people
- 13:38want to be using, then you have to
- 13:40decide for yourself which one makes more
- 13:41sense. For example, if you're a heavy
- 13:43user, you might benefit more from having
- 13:46an ultra plan at $310, as it's like
- 13:48twice as cheap as using the API here.
- 13:51Keep in mind, this is as of making this
- 13:53video, these prices might change. So,
- 13:56what is actually the benefit of the API
- 13:57and how do you start using this? Well,
- 14:00let me explain. What you need to do is
- 14:01you need to go over to the Higgsfield
- 14:03dashboard, then you click on API. This
- 14:05will open up a new tab, and here you can
- 14:07explore the API. You can then go to the
- 14:10quick start page right here, and you can
- 14:12copy this prompt and send this to your
- 14:14agent. If you copy this, you put it
- 14:16inside of your ChatGPT or your Claude or
- 14:19whatever you're using, you paste it in
- 14:21there, and then you hit send, and now it
- 14:23will start building your own AI video
- 14:26model tool for you using the Higgsfield
- 14:29API. Now, for me, this took like 10
- 14:31minutes, and it made basically our own
- 14:33variation of Higgsfield, but it is using
- 14:36the API. Like, it's exactly the same as
- 14:38Higgsfield. You have all of the models
- 14:39that you want to use, you have all of
- 14:41the image models that you can try out
- 14:42here, and you can add in your start
- 14:44frame, end frame, tweak these settings,
- 14:46everything you want. All you have to do
- 14:48here is you have to connect your API
- 14:50key, and then you can get this to work.
- 14:52Keep in mind that if you go to your
- 14:54Claude, Codex, or even Gemini chat where
- 14:57you start building this, that you can
- 14:59customize everything out of this. For
- 15:02example, I made a better type of
- 15:04Higgsfield, I would say, where I use
- 15:06this API access, and I customized it
- 15:09using all of my different workflows that
- 15:12me and my team are using internally. And
- 15:14I basically put this into the LLM, and
- 15:16then I've made variations of that, and
- 15:18that is this result. So, now there are a
- 15:21cool things which I can do. For example,
- 15:23if I added my prompt right here, I have
- 15:25this prompt optimizer and this prompt
- 15:27optimizer is made based on my own MD
- 15:31files. So I have had that in there. So
- 15:33it's using my own optimizer skill. It's
- 15:35not using anyone's else, completely
- 15:37tailored to my liking. So let's generate
- 15:39an example here. I'm just going to do
- 15:40the low quality mode. So then we hit
- 15:43generate and now it's using the API to
- 15:45make this. So this is the video that it
- 15:47made.
- 15:58And as you can see, it has used my MD
- 16:01file. I don't have to do this in my LM
- 16:03anymore because it's has done that for
- 16:06me through this new tool that I built.
- 16:08And you can access this right now. So
- 16:09the all you have to do is once you
- 16:11access this site, you just need to add
- 16:13in your own API key. You need to add an
- 16:15open AI API key. You need this so it can
- 16:18do the prompting for you and then you
- 16:19get access to this. Another cool feature
- 16:21that I built into this is the style
- 16:23feature and here you can copy the style
- 16:26of any type of reference image. So for
- 16:28example, I really like the look of this
- 16:31image and if I now drop this into the
- 16:33style right here, it will apply that
- 16:34style to my videos. So now I'm going to
- 16:36describe my scene. So I'm going to add
- 16:38in a simple prompt of a cyberpunk style
- 16:41samurai walking into a flower field and
- 16:43now we're going to hit generate and now
- 16:45it's using the style of this image and
- 16:47it's optimizing my prompt to generate
- 16:49something like this.
- 17:00So that is pretty cool. I've also got
- 17:02one more feature which I'm not going to
- 17:04share with you, but like this is a tool
- 17:06we use internally and that's why I like
- 17:09API so much because I can tweak my own
- 17:11MD files, my own skill files and I can
- 17:14just share it internally with my team,
- 17:15build my own tools, host it on my own
- 17:17servers, and pay only for what I use.
- 17:20So, if you apply all of these different
- 17:21methods, you're able to save yourself
- 17:24some credits. And hopefully, this was
- 17:26useful to you. If it was, please let me
- 17:28know in the comments down below what you
- 17:29want to see next. And if you want to
- 17:31learn how to make super realistic AI
- 17:33videos, then click the video that's on
- 17:35the screen right now, and hopefully, you
- 17:37will learn a lot more.
About this transcript
This page contains the full transcript of Stop Wasting Credits on AI Video Generators Like Seedance 2.5 by Dan Kieft, generated from the public captions YouTube serves with the video. The transcript has 3,087 words across 424 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.