YT YouTube2Text

YouTube2Text → Transcripts → Amazon FBA Product Research w AI 2026 (Claude Cowork)

Amazon FBA Product Research w AI 2026 (Claude Cowork) — Transcript

by Chris Rawlings · 4,733 words · 664 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00A year ago, Amazon product research this
0:02deep would have taken me about 11 hours
0:03all in, or I would have had to pay a
0:05consultant about a thousand dollars to
0:07get this done. This is actually wild,
0:11because with Claude, I just got a full
0:13product development and market analysis
0:15plan including specific product designs
0:17that meet holes in the market where
0:19there's demand but no supply, detailed
0:21and accurate financial projections,
0:23pricing and margin modeling, features to
0:26include based on customer feedback
0:27analysis, and even supplier quotes from
0:30an AI
0:33agent-based tool called Claude Co-work.
0:35Now, this is not the popcorn McDonald's
0:38AI that you're used to. This is not
0:40ChatGPT. This is an agent-based AI, not
0:44a chat-based AI. And there's a big
0:46difference between the two. Chat-based
0:48AI that you're probably already used to
0:49with ChatGPT and Grok and similar tools
0:51like that is like talking to an advisor
0:54or a friend. It's like a very
0:55knowledgeable friend, but it's more like
0:57a conversation. You say something, then
0:59it says something back to you. Then you
1:01say something again, and it says
1:02something back to you. Agent-based AI is
1:04not like that at all. Agent-based AI is
1:06more like an employee. You give it
1:09complex responsibilities or tasks, and
1:11it goes and executes them. Sometimes it
1:13takes time to execute them, sometimes
1:15even hours to execute them because it
1:17will execute complex tasks that involve
1:20a lot of different moving parts. It can
1:22involve research online, it can involve
1:24going to websites and filling out forms.
1:27And these complex tasks can involve
1:29actually multiple agents. So,
1:31agent-based AI with Claude Co-work, what
1:34it actually is is agent orchestration.
1:37It's kind of like asking someone to do
1:39something with AI, but you're asking AI
1:41to do something with AI. So, Claude
1:43Co-work is the orchestrator and that it
1:45enlists a bunch of agents to do
1:48different things. It can even use your
1:50Chrome browser in live time while you're
1:52doing something else in another Chrome
1:54tab, just like a person working next to
1:56you at your desk. So, while you're
1:58working, Claude can actually be working
2:00alongside you on a similar project or
2:02the same project as sort of your
2:03companion. Now, there are so many
2:05applications for this, and this is
2:07totally game-changing for Amazon sellers
2:09and any e-commerce brand owner. I'm
2:10going to be making a lot more videos
2:11about this coming up in the next couple
2:13of weeks, but in this video we're
2:15talking about one specific application,
2:17which is product research and
2:18development. So, what I did and that I'm
2:20going to walk you through in this video
2:22is I used Claude Co-work to come up with
2:24a full and complete plan and financial
2:27model and analysis of a product that I
2:30want to launch in a particular niche.
2:32And what it came out with was actually
2:34really incredibly valuable. It truly was
2:37something that you would have to pay a
2:38product development consultant or firm a
2:41lot of money to produce because it
2:43required sifting through a lot of data,
2:45coming up with really deep insights and
2:46analysis based on that data, and then
2:48making decisions and doing research
2:50based on it. So, it found suppliers for
2:52me. It found the holes in the niche
2:54where the market was underserved um
2:56because there wasn't a much supply, but
2:58there was a lot of demand based on the
2:59keyword data. In order to do something
3:00like this, you need to know how to use
3:03Claude Co-work right, and I am going to
3:05be coming out with a video on the
3:06fundamentals of how to set this up, but
3:07you can actually get started right away
3:09without a super involved setup. All you
3:11have to do is download Claude AI for
3:14your desktop, then on the top you'll see
3:16a little switch that says Co-work. You
3:18move over to Co-work, and now you're
3:20using agentic AI where Claude is an
3:22orchestrator of multiple agents that
3:25work on your behalf. Now, you can
3:26connect things to Claude Co-work so that
3:28it can work on your behalf in particular
3:30tools that you live in like Slack, for
3:32example, it can start sending messages
3:34on your behalf, or Notion if you wanted
3:36to have access to your entire SOP
3:38library and be able to read it and
3:39update it in live time. There are a lot
3:41of connectors. We're not going to go
3:42through all that in this video. We're
3:43going to just do a quick start with this
3:45very impressive application, which is
3:48product research and development. So,
3:50first let's go through the results here,
3:52and then I'm going to walk through how
3:53we get the result, and it's actually a
3:54lot easier than you'd expect. But, you
3:57have to know how to do it right. With
4:00agentic AI, it's really all about
4:02context. Unlike a chat-based AI where I
4:04would just type in a little prompt or a
4:06question and get a response, when I'm
4:08giving essentially, you have to think
4:10about it like giving a person a job.
4:12When I'm giving this agentic AI a job, I
4:14need to give it the proper context and
4:16resources in order for it to do the job
4:18properly. So, for this project, I chose
4:20the niche of backseat covers for dogs
4:23just as an example to see what we could
4:26do in this space. We actually have a
4:27brand in our portfolio that already has
4:29products like this, and I just wanted to
4:30see what it would come up with so I
4:32could compare it to product that I'm
4:34already very familiar with. And what I
4:37got back was much more incredibly
4:40valuable than I really ever could have
4:42imagined, which was this very succinct
4:46but detailed report on if I'm going to
4:49develop another product in this space,
4:51what to focus on based on the actual
4:53opportunity there, along with financial
4:55projections, unit economics, supplier
4:58quotes from Alibaba, and a bunch of
5:00other really valuable stuff that would
5:02have cost me either over a thousand
5:03dollars or over 10 hours of my own time
5:06if I hadn't done this with Co-work. And
5:08with Co-work, I think it took a total of
5:11maybe about six or seven minutes,
5:13something like that in total for all of
5:15the the work that I had to do to put
5:17this together. [music] And take a look
5:18what we got here. I mean, we literally
5:21have market gaps where it's specifically
5:24telling me this is where there's demand
5:26but low supply based on analysis. Now,
5:29here's where I want to start telling you
5:30about the context that I gave it because
5:32this is really important. In order for
5:34Claude Co-work to do this job properly,
5:37I exported a big file of keyword
5:40research data from Helium 10. Now, this
5:43literally took me only two minutes. I
5:44just logged into Helium 10, which is a
5:46keyword research tool for Amazon
5:48sellers. I typed in dog back car seat
5:50bed, and I exported all the data. I
5:52didn't filter it at all. I didn't It was
5:54just a raw CSV file that I exported. I
5:57imported it into Claude. I knew that it
5:59would be able to do the data analysis,
6:01and it did it very well. Claude is very
6:04[music] good, much better at any of the
6:06other models, ChatGPT or any of the
6:07other ones, at data analysis. It's
6:10really, really good. In fact, it has a
6:11native plugin for Excel so that inside
6:15Excel you can actually work with Claude
6:16alongside it and have it manipulate your
6:18data in your spreadsheet as you're
6:20working. So, I gave it that context
6:22because that told it all of the insider
6:25data it needed to know about the search
6:26demand and the keyword data, which is
6:28highly related to the the actual demand
6:31in the market and the competitive data
6:32for the space. And then the rest of the
6:34research it did on its own through
6:35controlling my browser and through doing
6:37its own back-end research directly. And
6:39so, look at these insights. These are
6:41insights that I would have had to sift
6:43through hours of keyword research data
6:46and competitor reviews and all kinds of
6:48other stuff to really get these insights
6:49myself, but it found these market gaps.
6:51But I looked through the keyword
6:52research data myself just to validate
6:54this, and it is indeed true that these
6:56would be good products to launch. They
6:58have significant amounts of search
7:00volume across all the keywords that
7:02speak to these particular shopper
7:04intents, but they're really underserved
7:06on the market, and there's really not
7:08that many good offers for it. So, this
7:11insight is really critical. I mean, this
7:14is really the core decision when
7:16launching any new product is is this
7:18product going to be successful? Is it
7:20really speaking to a big market demand?
7:22Do I have some kind of unique spin that
7:23I can put on my product that gives me an
7:26edge? From the very beginning, we have
7:27that based on this analysis, but this is
7:29just the tip of the iceberg. So, let's
7:31scroll down here. Here we go a level
7:32deeper with keyword categories by search
7:34volume where we can see what underserved
7:37areas of the market by feature or by use
7:40case. So, we have C still here, large
7:43dogs vehicle specific, but we also have
7:45safety restraint. So, like the hookup to
7:47keep the dog safe while it's in the
7:48backseat. And then luxury premium is in
7:51here as well. Then we have some that are
7:52partially served like small dogs, for
7:54example. And then we have some that are
7:56well served like the hammock style. It
7:58correctly brought up the insight that
8:00the hammock style portion of the market
8:02is actually overserved. A lot of them
8:04focus on that. Probably means there's a
8:06lot of demand as well, but it's very
8:08overserved or well served in the market.
8:11So, that wouldn't be a differentiating
8:13factor for us, and it correctly pointed
8:15that out. Now, stick with me. This gets
8:16really crazy like further down where it
8:18actually goes and gets supplier quotes
8:21and adds them in here. And by the way, I
8:22checked these, and these are also
8:24accurate. And breakdown of the unit
8:27economics and margin modeling and all of
8:30that, but we're not there yet. Then we
8:31have the market analysis of the current
8:33landscape. Now, this is really valuable
8:35as well. So, we see some of the
8:37different categories that are currently
8:39all out there like hard bottom, elevated
8:42um booster seats. This is to make dogs
8:45higher up so they could see through the
8:46window, full coverage that go all the
8:48way across including doors, uh cargo
8:50trunk covers, convertible or
8:53multi-function ones. There's bench style
8:55ones, and it separated it out into these
8:58different categories so that I could
9:00know where the one that I'm thinking of
9:02doing would fit into this. Then we got
9:04into the price tiers so we could see
9:06where we'll be sitting on the scale.
9:08There's budget range of 20 to 50
9:10dollars, the mid-range of 50 to 100
9:12dollars, and then super premium ones at
9:14100 to 200 dollars. Now, here's another
9:16super valuable part. We get into top
9:18customer complaints. Through reviewing
9:21all of the reviews, Co-work was able to
9:23find what the biggest issues are with
9:26the product so that I can solve them
9:28from the get-go and use those as unique
9:29selling propositions and avoid bad
9:31reviews myself. So, it identified that
9:34seam failure, which is just an actual
9:36product defect or a product quality
9:38issue, is one of the main issues. That
9:40there's waterproofing failures, which of
9:42course is important with dogs cuz they
9:43pee and they're they step in mud and
9:45stuff. There's fit problems so I can
9:47make sure to include the dimensions in
9:49the image when I launch this product.
9:51There's lack of instructions, that's a
9:52super easy one to solve, just very
9:54visual, obvious instructions. So, all of
9:56these things are things I'm going to
9:58take into consideration when developing
10:00the product and working with the
10:01supplier and the sourcing agent to get
10:04this made in a way that I know has a
10:06high likelihood of success. And then
10:07here we dive deeper into the market gap
10:09analysis. So, we have the different
10:11segments, the large dog, which I it was
10:13one I totally agree with, the vehicle
10:15specific designs, which really mean
10:18truck and SUV, like this state specific
10:20vehicles, but what it came down to when
10:22I looked into the keyword data was
10:24actually like talking either truck or
10:26SUV. Those are the the that's the data
10:28that comes up a lot. But, you could even
10:30go as deep as having child variations
10:32that speak to specific cars that are
10:34really popular, like a Tesla, for
10:36example. The temperature regulation, I
10:38think is a good point, but I looked into
10:40it from a product design perspective, I
10:41think it would be impractical. So, I
10:43think this is a good insight, but it's
10:45not practical to actually implement. The
10:46senior dog is also possible, and that is
10:49a pretty big market. That's something
10:51that I would consider. And the Spanish
10:53language market, I think that's an an
10:55interesting insight, but it doesn't
10:58really affect the product development.
11:00So, that's more something that would
11:01come into play when it came into our
11:03advertising. Now, it ranked these unique
11:05sub niches for us, and I'd say that I
11:08don't really agree with the ranking
11:10completely. I would say the large dog
11:11and truck for sure, and then I would put
11:13senior dog up here, and then Spanish and
11:15temperature would be at the bottom. But,
11:16still it's it's actually pretty close to
11:18what I would agree with and what I think
11:19I would have probably come up with after
11:21a long time doing product research for
11:24developing this product myself before
11:26initiating any kind of deal with a
11:27supplier. So, it doesn't stop there.
11:29Then it makes its own decision, and I
11:31can choose to agree or disagree with
11:33this, of course, but it made the
11:35decision that this would be the most
11:38effective sub niche of this niche to
11:41launch, which is large dog premium hard
11:45bottom seat covers. And it got to this
11:48conclusion through actual data analysis.
11:51Now, I know I haven't even shown you
11:52guys how to generate this yet, and I'm
11:54going to get into that in just a second.
11:55But, we're just going to finish going
11:56through this first, and then I'll show
11:57you exactly how to do it. So, then we
11:59start getting into the modeling of how
12:00to actually make it happen and what the
12:01economics will look like once we make it
12:03happen. So, it went and found quotes
12:05from specific suppliers on Alibaba with
12:09their freight on board price for this
12:12particular product that it came up with
12:14that solves this particular unmet need
12:17in this particular niche. Pretty cool.
12:19Now, I went on Alibaba to validate these
12:22prices, and they are in fact in range.
12:24It even gave us a breakdown of the
12:26different components of the cost. So, if
12:28we wanted to tweak our cost by removing
12:31or adding a different feature, for
12:33example, the hardware or the non-slip
12:35bottom, we could see how that would
12:37affect the cost. Now, this is insight
12:39that you would not be able to get
12:40yourself unless you are grilling the
12:42manufacturer, and even then a lot of
12:44times they're very shy to kind of fork
12:46up this information. It then did
12:48financial modeling for us so that we
12:50know what our per unit margin is going
12:51to be. Now, this I think it was a little
12:54bit too liberal with this. I would have
12:56been more conservative. I think the COGS
12:57is going to be higher for the premium
12:59tier of this product, for example. Uh it
13:01took into consideration the shipping
13:04fee, but I think it was under quoted,
13:06and it also didn't include land delivery
13:09uh in the US, and it didn't include the
13:11fee that you pay Amazon to receive the
13:13delivery or any kind of inventory
13:15storage fee. So, there was some stuff
13:16missing here, but it got I would say it
13:19got it like 70 to 80% right, and it was
13:22a good estimate uh in general to kind of
13:24get an idea of where you're at, and it's
13:26starting place so that you can then just
13:27tweak it. And this is the case a lot of
13:29the times with AI where it gets you like
13:31most of the way there, and then you
13:32really just have to check and alter it
13:34to get it the last like 10 to 20% of all
13:37the way home to make it really
13:38practical. But, this was a really good
13:41starting place. It did a model of the
13:42total initial investment required,
13:45including the first production run, the
13:48custom setup and molding fee, the
13:50engineering fee, product photography,
13:52which you could avoid by using AI, of
13:54course, if you really wanted to, and the
13:57PPC that you would have to pay. Now, I
13:59again would say that to actually get
14:01this product profitable, it's going to
14:03take you longer than this report is
14:07assuming. So, I would be again a little
14:09bit more conservative with this, meaning
14:11I would probably end up with a budget
14:13higher than this, but I don't think it's
14:15too far off. In all honesty, I think
14:17it's up it's a pretty good initial it's
14:20in the ballpark. It's in the ballpark,
14:22and I would I would tweak this a bit to
14:24come up with my my full budget for
14:26launch for this product, and I
14:28personally would be more conservative
14:29with how long it would take to get to
14:31profitability of it, to get enough
14:32reviews, and to get enough sales
14:33velocity to start ranking. But, still it
14:36wouldn't be so far It's not like in
14:38another dimension. It's not going to be
14:40like 10 times this. It's just going to
14:41be maybe a little higher than this. Then
14:43it even created a plan of action. So, it
14:46talked through exactly what to do to
14:49contact the suppliers, what steps to
14:51take with the supplier to do the product
14:53development and all the customization,
14:55how many units to order. Now, this I
14:57would also disagree with this. I tend to
14:59order more units up front just so that I
15:01have enough to push really hard for
15:03ranking. But, these are more expensive
15:05products per unit, so this may actually
15:08be a realistic unit level depending on
15:11the level of the seller who's actually
15:12launching this. But, again, this is a
15:14really great starting place, and a lot
15:16of this data, especially the keyword
15:18data and the market niche analysis and
15:20the gaps in the market, this is stuff
15:22that it would have taken a long time to
15:25really get a handle on. Because you
15:27might observe gaps in the market
15:29yourself by doing keyword research data
15:31and market analysis, but you're not
15:33going to have all of the data all at
15:35once in your head. So, you might be
15:37biased for the keywords that you see
15:38first. You might have seen some keywords
15:40about senior dogs, for example, and then
15:42you get obsessed with that, and you
15:43think that's a really great idea,
15:45whereas that might have been an
15:46opportunity, but there's another one
15:47that's a much bigger opportunity, like
15:49in our case large dogs, or it's the dog
15:51beds for specific cars, like trucks or
15:54SUVs. Now, let's get into how I actually
15:57generated this report, and you're going
15:59to find it fairly simple and very easy
16:03to do. I didn't have to use a bunch of
16:04different connectors, and I'm going to
16:06do a whole video on how to set up Claude
16:09CoWork properly for an Amazon seller in
16:11the future. But, for now we're just
16:12keeping it really simple. All I did to
16:15generate this very detailed useful
16:18product development research report was
16:20a single prompt, and I uploaded the
16:23keyword research report from Helium 10
16:25for the product. So, I said I'm
16:26launching a new product on Amazon, it'll
16:28be a dog back car seat cover for a car.
16:30I need you to research all beds that are
16:32available on Amazon, then
16:33cross-reference it with the keyword
16:34research analysis I'm providing from
16:36Helium 10, then find holes in the market
16:38where there's a keyword research for a
16:39particular shopper intent that's not
16:40being filled in the market currently.
16:42Then come up with a product that would
16:43fill that need, find suppliers for the
16:45product for the best possible price. You
16:47can put all of this information into one
16:49report for me to review. That was the
16:50full prompt. I'll copy it in the
16:52description of this video if you want to
16:54paste it in and try it out yourself,
16:56which I highly recommend. You can watch
16:57all the videos you want on AI, but if
16:59you don't just start playing with it and
17:01actually utilizing it, you're never
17:03going to actually learn it. Watch it.
17:05You The breakdown is usually spend 90%
17:08of your time on learning a new thing by
17:10actually doing that new thing, and 10%
17:12of your time watching videos or reading
17:14books or learning about that new thing.
17:16If you want to surf, you got to spend
17:18all your time in the water surfing, not
17:20all your time reading books about
17:21surfing. That's just how it works. So,
17:23I'd recommend if you're watching this
17:24video, do this now. Just download Claude
17:27on your desktop and try this exact same
17:29prompt, but do it for your product
17:31category. And what I did with this was I
17:33uploaded a keyword research report from
17:36Helium 10. For those of you guys not
17:38familiar with Helium 10, Helium 10 is a
17:40tool for Amazon sellers to do keyword
17:43research and market analysis and keyword
17:46rank tracking and things like that. So,
17:47the tool within Helium 10 is called
17:49Cerebro. Helium 10 has free trials, by
17:52the way, and there are also other tools
17:53that do this. You don't have to use
17:54Helium 10. There are lots of tools that
17:56do keyword research for Amazon sales.
17:58Many of them are free or have free
17:59trials, so I'm not married to Helium 10.
18:01You could use any tool you want, but I
18:03used Helium 10, and all I literally did
18:06was go to Cerebro, type in dog back seat
18:08cover. I said get keywords, and then it
18:10produced this list, and the list was a
18:12lot. It was like 11,000 keywords, and
18:15you could see if I order it by keyword
18:17search volume, there was actually a lot
18:19of keywords in here that were not
18:20relevant, like iPhone 17 Pro Max, and uh
18:24car accessories is is not super
18:27relevant, but it would fit in there. Dog
18:28bed hammock, um a lot of irrelevant
18:31stuff. Uh and that's because I ordered
18:33by search volume, but the point is I
18:34didn't do any filtering or vetting of
18:36this whatsoever. I just let Claude do it
18:39all, and Claude has the context and the
18:42ability to analyze data such that it
18:44doesn't need it to be clean. I literally
18:46just clicked export data to a CSV file,
18:50downloaded that, and then I uploaded it
18:52to Claude with my prompt. And this was
18:55the key to how I got actually good
18:57results from this. When I uploaded the
18:58proper context and I had a detailed
19:01prompt, I got the information that I
19:04need from it, and it generated this
19:06report for me, which turned out to be
19:08fairly accurate and very useful. Now,
19:10one thing you do have to do is set up
19:13your desktop folder, because Claude
19:16CoWork, unlike chat-based AI, it
19:19actually runs on your desktop. It's
19:21using your desktop hardware. It's not
19:23using cloud hardware. It's using your
19:26hardware. So, it's running on your
19:27computer. Like if you shut your computer
19:29down, it would stop the task. And now it
19:31doesn't end there. Once I got that
19:33report and I made some decisions about
19:35it about what product I want to launch,
19:37I'm going to continue the entire process
19:39with Claude CoWork now. This is just how
19:42to run an e-commerce business in 2026
19:44now. Things have changed so dramatically
19:47so fast. So, what I did was I asked what
19:49are all of the keywords related to large
19:52dogs and their associated search volume
19:53so that I could start putting together
19:56my campaigns, my PPC campaigns for when
19:59my product comes into play. I'm going to
20:01do the same thing for every element of
20:03this product development and launch is
20:05work with Claude alongside me while I'm
20:08doing the product launch. And this type
20:10of co-working alongside AI is how we're
20:13starting to do everything inside our
20:15portfolio. And this is what every Amazon
20:17seller is going to have to do. You're
20:18seeing stories of left and right of
20:2017-year-olds starting e-commerce brands
20:23that they own they have no employees and
20:24they just use AI and then they rocket up
20:27super fast and they're already competing
20:28with years-long entrenched competitors.
20:30This is happening now. And if you're
20:32still doing things the old way, you're
20:34just choosing to do things in a way that
20:36takes 10 times more time or 100 times
20:38more time than somebody else who has
20:40adopted these tools. So, if you're a
20:42person who's been kind of putting this
20:43off, I recommend you download and start
20:45trying it. I'm going to be coming out
20:47with some more advanced videos in the
20:48next couple of weeks with a lot of the
20:50other things that we're doing with
20:52Claude co-work and with agent-based AI
20:54inside e-commerce brands. So, look out
20:56for that. And if you want this
20:57particular report that I was just going
20:59through so that you could upload it to
21:00Claude and have it generate the same
21:02report for you, I put a link to download
21:04the report in the description of this
21:06video so you could tap the link there.
21:07It won't stop this video. You could tap
21:09it now, download the exact report I got,
21:11upload it to Claude co-work and ask it
21:13to create the exact same one for you.
21:16Now, here's where we get meta. When I
21:18went about making this YouTube video,
21:20just for shits and giggles, I asked
21:22Claude to write the script for me. And
21:24you can see
21:26as you could see in the beginning of the
21:28script, [music] it's actually pretty
21:29similar to how I started this video.
21:31Now, does that make me just a meat
21:33puppet for AI at this point? Hell no.
21:36The robot overlords have not gotten me
21:38yet. I ended up just riffing for this
21:40video because I don't like following
21:41scripts anyway and I didn't agree with a
21:43lot of the stuff that it said. But I
21:44just wanted to see what it would come up
21:46with and it actually was quite good. Had
21:48some good points and I will some and the
21:51beginning hook that I had for this video
21:53was similar to what it came up with. So,
21:55it's really crazy. I think this is going
21:57to be a year that just changes
21:59everything about how we work as
22:01entrepreneurs, especially in the
22:02e-commerce space and everyone will have
22:05to adapt this one time or another. So,
22:08if you haven't done it now, this is a
22:10very easy, simple first task that you
22:13can do to start using Claude co-work and
22:15agent-based AI in your business.
22:17>> [music]
22:17>> I recommend you try this particular
22:19exercise that we just did. And if you're
22:20really interested in utilizing AI to run
22:23Amazon brands, check out this video
22:25where I walk through how to use Google's
22:27AI image generator to generate
22:29production quality, high conversion rate
22:32and high click-through rate images for
22:34your product essentially for free or
22:36near free that would have cost hundreds
22:38or thousands of dollars just last year.
22:41So, check out that video here and I'll
22:42see you over there.

About this transcript

This page contains the full transcript of Amazon FBA Product Research w AI 2026 (Claude Cowork) by Chris Rawlings, generated from the public captions YouTube serves with the video. The transcript has 4,733 words across 664 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.