YouTube2Text

How To Learn To Code In 2026 — Transcript

by Tina Huang · 3,655 words · 531 segments · language en · Watch on YouTube

Full transcript

  1. 0:00Learning to code is so different in
  2. 0:012026. So, in this video, I shall explain
  3. 0:04to you how to learn to code in 2026.
  4. 0:07Starting off with what coding actually
  5. 0:08looks like today, it looks very
  6. 0:09different. Then I'll explain what you
  7. 0:10should be learning that is still
  8. 0:11relevant right now and for the future.
  9. 0:13And finally, some tips on how to
  10. 0:15supercharge your learning. Oh, by the
  11. 0:16way, just in case you have any doubt
  12. 0:17about whether you should still learn how
  13. 0:19to code in 2026, the answer is yes.
  14. 0:21There are still so many opportunities
  15. 0:23that are only available to people who
  16. 0:25know how to code. Plus, learning to code
  17. 0:27in 2026 is so much easier and faster
  18. 0:30than it used to be. So, it's kind of a
  19. 0:31no-brainer if you want to build stuff.
  20. 0:33Now, with that out of the way, in case
  21. 0:34you don't know me, hi. My name is Tina.
  22. 0:36I used to be a data scientist at Meta
  23. 0:38and I have been coding for 10 years.
  24. 0:40Without further ado, let's go. A portion
  25. 0:43of this video is sponsored by Warp. On
  26. 0:45February 3rd, 2025, Andrej Karpathy, the
  27. 0:47previous director of AI at Tesla and the
  28. 0:49founding member of OpenAI, made a viral
  29. 0:52tweet defining a new term called vibe
  30. 0:54coding. He explains, "There's a new kind
  31. 0:55of coding I call vibe coding where you
  32. 0:57fully give in to the vibes, embrace
  33. 0:58exponentials, and I forget that the code
  34. 1:00even exists. It's not too bad for
  35. 1:02throwaway weekend projects, but still
  36. 1:04quite amusing." Then, almost exactly a
  37. 1:06year later, on February 5th, 2026, he
  38. 1:09made another post. "Today, 1 year later,
  39. 1:11programming via LLM agents is
  40. 1:12increasingly becoming a default workflow
  41. 1:14for professionals, except with more
  42. 1:15oversight and scrutiny." Now, to
  43. 1:17differentiate this much more
  44. 1:18sophisticated integral workflow from
  45. 1:21vibe coding, he said his current
  46. 1:22favorite term for it is agentic
  47. 1:24engineering. You have your terminal/IDE
  48. 1:26open where you have multiple coding
  49. 1:28agents coding furiously away for you and
  50. 1:30you're like, "Build this. Fix this. No,
  51. 1:32don't do that. Do this instead."
  52. 1:34Orchestrating your AI agents like a
  53. 1:36puppet master. But, do not be fooled. It
  54. 1:39may look like you're just speaking
  55. 1:40English to the AI agents and just like
  56. 1:42orchestrating them, but you need to be
  57. 1:44an experienced software engineer to even
  58. 1:46do that. You need to know how to
  59. 1:47structure what it is that you want to
  60. 1:48build in a way that the AI agents can
  61. 1:50understand, be able to provide context
  62. 1:51and documentation. And in order to
  63. 1:53correct the AI and tell it that it's
  64. 1:55wrong, you need to actually know what is
  65. 1:56right. To be a good manager, you need to
  66. 1:58know what it is that you're managing.
  67. 2:00So, that is why if you want to learn how
  68. 2:02to code in 2026, you still need to know
  69. 2:04the fundamentals of coding and
  70. 2:05engineering. But, instead of focusing so
  71. 2:07much on the details of the code
  72. 2:09implementation itself, it's about
  73. 2:11understanding big picture software
  74. 2:12design, systems, and architecture. And
  75. 2:15particular emphasis on security and
  76. 2:17privacy, because these are the things
  77. 2:18that AI coding agents tend to struggle
  78. 2:20the most with. Then, you can actually
  79. 2:21move on to take advantage of the AI
  80. 2:23coding tools that we have available
  81. 2:25today and really do agentic engineering.
  82. 2:27So, should I tell you what it is that
  83. 2:28you need to learn? All right, I'm going
  84. 2:29to list out the topics in a roughly
  85. 2:31chronological order that I think you
  86. 2:32should be learning, assuming you don't
  87. 2:34know anything. And also suggest some
  88. 2:35resources on these topics.
  89. 2:38Okay, so first thing that you want to
  90. 2:40learn is the basics of coding itself.
  91. 2:42Stuff like variables, types, [music] if
  92. 2:44statements, loops, object-oriented
  93. 2:46programming, and APIs. Also, I'm using
  94. 2:49Python as the demo language to learn,
  95. 2:51which is particularly good if you want
  96. 2:52to be building AI agents and doing data
  97. 2:54stuff. But, if you want to be doing
  98. 2:55something more like web stuff, like
  99. 2:56developing web apps, you might want to
  100. 2:58start off with a language like
  101. 2:58JavaScript. It's okay though, because
  102. 3:00I'm going to give you guys a prompt at
  103. 3:01the end of the section that you can use
  104. 3:02to customize this learning plan specific
  105. 3:04to the language of what you want to be
  106. 3:06learning. Okay, great. So, these are the
  107. 3:07basic basics. Like, you really need to
  108. 3:09know these, otherwise you wouldn't even
  109. 3:11be able to read the code that the AI
  110. 3:13generates.
  111. 3:13>> That sign can't stop me, because I can't
  112. 3:16read.
  113. 3:17>> And I'm going to put on screen now some
  114. 3:18of my favorite resources on this topic.
  115. 3:20I personally still recommend going with
  116. 3:21these human-made resources, uh because
  117. 3:24we know that they're really good.
  118. 3:25They've been vetted by thousands of
  119. 3:26engineers. Besides, all AI-generated
  120. 3:28code is based upon code that was written
  121. 3:30by human developers for decades. Next
  122. 3:32up, you need to learn about software
  123. 3:33architecture. What that means is how is
  124. 3:35it that projects are structured? How do
  125. 3:37you choose a tech stack? The system
  126. 3:39design. What kind of APIs should you be
  127. 3:40using? And how the data flows through
  128. 3:42your software, and where is it being
  129. 3:43stored? What kind of databases are you
  130. 3:45using? And testing. Testing is so
  131. 3:46important, especially if you're using AI
  132. 3:48coding agents. You need to understand
  133. 3:50how to write test cases and how to test
  134. 3:51different of your software. And finally,
  135. 3:53how does deployment work? Where are you
  136. 3:54going to be hosting your software? This
  137. 3:56is like the bread and butter of software
  138. 3:58engineering and you really need to
  139. 4:00understand these high-level things to be
  140. 4:02able to direct your AI to be building
  141. 4:05the things that I should be building and
  142. 4:06redirecting it when it's doing something
  143. 4:07wrong. You need to be able to
  144. 4:09conceptualize what the end product
  145. 4:10should look like based upon the
  146. 4:12requirements that you have. For example,
  147. 4:14at work, you might want to build a
  148. 4:15software um that stores a massive amount
  149. 4:17of information and have lots of people
  150. 4:18querying it all the time. So, then you
  151. 4:20need to be thinking, what sort of
  152. 4:21database should I be using? That is
  153. 4:22cheap, safe, and fast. And how do I
  154. 4:24design the user interface that allows
  155. 4:26for multiple people querying at the same
  156. 4:27time? These are all like design
  157. 4:28questions that you need to be able to
  158. 4:30understand. I'm going to put on screen
  159. 4:31now some resources that you can check
  160. 4:33out to learn about these topics. Next up
  161. 4:34is version control and GitHub. By
  162. 4:36version control, we mean a system that's
  163. 4:38able to track the changes in your files
  164. 4:40and code over time. So, you can track
  165. 4:42the history, revert back if you need to,
  166. 4:44and collaborate simultaneously with
  167. 4:46other people as well. There are actually
  168. 4:47a lot of different types of version
  169. 4:48control software out there. Um the most
  170. 4:50popular one is called Git. So, people
  171. 4:52would use Git in order to do version
  172. 4:53control and then put their code on
  173. 4:55something called GitHub. GitHub is the
  174. 4:57industry standard for displaying your
  175. 4:59code, sharing it, and collaborating with
  176. 5:01other people. This is especially
  177. 5:02important if you're intending to do
  178. 5:03agentic engineering because you want to
  179. 5:04be tracking the changes that your AI
  180. 5:06coding agent is doing, right? And in
  181. 5:07case it makes a mistake, you can revert
  182. 5:09it and redirect it without like losing
  183. 5:11your entire code base. Which by the way,
  184. 5:13this happened to a lot of vibe coders
  185. 5:15who did not know the fundamentals of
  186. 5:16coding and engineering. I'll put on
  187. 5:18screen now some resources for you to
  188. 5:19learn about version control and GitHub.
  189. 5:21All right, next up is security and
  190. 5:23privacy. So, this is a topic that was
  191. 5:25not particularly important when somebody
  192. 5:28is learning how to code um like a few
  193. 5:30years back. Like obviously security and
  194. 5:32privacy is important, right? But it's
  195. 5:33sort of something that just gets taught
  196. 5:34like within the scope of other things.
  197. 5:37Like how does authentication work and
  198. 5:38how do you like integrate authentication
  199. 5:40into your apps? But the reason I'm
  200. 5:41separating this out as its own like
  201. 5:43little topic here is because I I it's
  202. 5:45particularly important in this day and
  203. 5:47age to learn about security and privacy
  204. 5:49because it's usually one of the blind
  205. 5:50spots of AI coding agents. So you as the
  206. 5:52human can't just kind of like hope that
  207. 5:54the AI agent is just going to take care
  208. 5:56of it. You do need to be quite explicit
  209. 5:57in incorporating these principles. Don't
  210. 5:59worry, this is honestly not like a
  211. 6:00really big topic. It's just really
  212. 6:02important to know. And as they say, I
  213. 6:05think, paranoia makes a good engineer,
  214. 6:07maybe. I'm going to put on screen now a
  215. 6:09few different resources that I suggest
  216. 6:10you take a look at and learn about
  217. 6:12privacy and security. All right, so at
  218. 6:13this point you would have covered pretty
  219. 6:15much the fundamentals of coding and
  220. 6:17software engineering now. You have now
  221. 6:19earned your stripes to be able to
  222. 6:21harness the power of agentic engineering
  223. 6:25by using AI coding agents. Yay! But with
  224. 6:28that being said, I do want to throw in a
  225. 6:30bonus little topic, which is debatable
  226. 6:33whether it's actually like necessary
  227. 6:35necessary for you to know this topic,
  228. 6:37but I personally think it's pretty
  229. 6:38important and you're probably going to
  230. 6:39have to learn about it anyway. And that
  231. 6:40is microservices, otherwise known as the
  232. 6:42concept of containerization. This is a
  233. 6:44software deployment method that packages
  234. 6:46an application's code together with all
  235. 6:49of the files, libraries, and
  236. 6:50dependencies that it needs in order to
  237. 6:51run that piece of code. So it's sort of
  238. 6:53like little packages, kind of little
  239. 6:54containers of pieces of software. The
  240. 6:56purpose of this is to isolate the
  241. 6:58application from its surroundings so
  242. 6:59it's not interacting with the rest of
  243. 7:01your computer, um and it's able to run
  244. 7:03consistently across different computer
  245. 7:04environments, like different computers,
  246. 7:06different computer systems, different
  247. 7:07types of clouds. This is usually more of
  248. 7:09an advanced topic and people usually
  249. 7:10learn about this when they're building
  250. 7:12stuff or like on the job. But the reason
  251. 7:13why I'm putting this here as like an
  252. 7:15optional but maybe you should learn
  253. 7:17about this kind of topic is as another
  254. 7:19precaution if you're using AI coding
  255. 7:21agents. You can imagine if you're
  256. 7:22letting your coding agent or even like
  257. 7:23multiple coding agents just do whatever
  258. 7:25it wants, right? It can end up messing
  259. 7:26up your development environment entirely
  260. 7:28or like building things in a way that
  261. 7:30cannot be properly deployed and used by
  262. 7:32other people. And you'll be just left
  263. 7:33like really frustrated cuz you can't
  264. 7:35figure out why it's happening. So if you
  265. 7:37do take the precaution to containerize,
  266. 7:39you're going to end up with something
  267. 7:39more reliable. And if your AI agent does
  268. 7:42run amok and starts doing crazy things,
  269. 7:45it's going to be in its own little
  270. 7:47environment doing it. So, worst case
  271. 7:48scenario, just shut off that
  272. 7:49environment. Let me know in the comments
  273. 7:51of these topics I have just covered, how
  274. 7:53many of these topics do you already
  275. 7:54know? Which ones do you still need to
  276. 7:56learn? Curious where you guys at on the
  277. 7:57learning journey. As promised, I'm also
  278. 7:59going to put a prompt on screen. Also,
  279. 8:01I'll probably link in the description,
  280. 8:02too. You what you can put into your
  281. 8:03favorite AI chatbot, ChatGPT, Gemini,
  282. 8:05Claude, Quen, Deep Seek, whatever you
  283. 8:07want, and it will help adapt this
  284. 8:09learning plan and resource list to fit
  285. 8:11the goals of your learning and also help
  286. 8:13you determine which coding language
  287. 8:14makes the most sense for you to learn.
  288. 8:16Yay, amazing. You have now earned your
  289. 8:18stripes to be able to tackle this new
  290. 8:21world, this brave new world of agentic
  291. 8:23engineering.
  292. 8:25Now, agentic engineering itself, gosh,
  293. 8:28it does agentic engineering just not
  294. 8:29sound as cool as vibe coding. It's just
  295. 8:31not as catchy.
  296. 8:33>> [gasps]
  297. 8:33>> It's fine. Agentic engineering, okay.
  298. 8:36Agentic engineering itself is also a
  299. 8:38skill. It involves how to set up your
  300. 8:40project and how to set up your AI coding
  301. 8:41agents, how to communicate what it is
  302. 8:43that you wanted to build to your coding
  303. 8:45agents, and how to monitor them
  304. 8:46correctly among many other things. For
  305. 8:48agentic engineering, there are different
  306. 8:50options for which coding agents to use
  307. 8:52and how to use them and orchestrate them
  308. 8:53together. Me personally, I really like
  309. 8:55using Warp, who is also the sponsor of
  310. 8:57this portion of the video. Because it's
  311. 8:59a pro developer tool that covers the
  312. 9:01full spectrum of AI coding. Getting
  313. 9:03started is so simple. I built this
  314. 9:05entire project using Warp. But, as it
  315. 9:07scaled, I needed to move faster, and
  316. 9:09that's where the magic begins. Instead
  317. 9:11of tackling one thing at a time, I
  318. 9:12deployed multiple coding agents
  319. 9:14simultaneously. Here's one that's adding
  320. 9:16a back-end feature, one on
  321. 9:17documentation, and one that's writing
  322. 9:18tests. This is all powered by Oz, which
  323. 9:21is a cloud-based agent orchestration
  324. 9:23platform built on top of Warp. The Oz
  325. 9:24dashboard gives you a real-time view of
  326. 9:27all your agents. At a glance, you can
  327. 9:28see what they're working on, their
  328. 9:29status, and any issues that they may
  329. 9:31have. And because it's cloud-based, each
  330. 9:33agent runs in its own isolated
  331. 9:35environment, meaning that they won't
  332. 9:36interfere with each other and my poor
  333. 9:38MacBook Air can stay alive. That is not
  334. 9:41all. You can actually incorporate Oz
  335. 9:43within your project itself. For example,
  336. 9:45this is a scheduled agent that updates
  337. 9:46my microsite every day. It goes in
  338. 9:47there, takes the AI news of that day,
  339. 9:49which is what I care about, and
  340. 9:50summarizes it and displays that
  341. 9:51information for me. With Oz, you can
  342. 9:53also convert skills into AI coding
  343. 9:55agents, integrate them directly with
  344. 9:56Slack and Linear, and tap into a ton of
  345. 9:58other features that support team
  346. 9:59workflows. 700,000 engineers are already
  347. 10:02using it and 97% of code diffs get
  348. 10:04accepted. Which, honestly, is pretty
  349. 10:06impressive and tracks from my
  350. 10:07experience, too. So, if you want to get
  351. 10:09started, you can get started for free at
  352. 10:10the link linked below. Thank you so much
  353. 10:11to sponsoring this portion of the video.
  354. 10:13Now, back to the video. Unfortunately,
  355. 10:15there are not that many good resources
  356. 10:17on agentic engineering, probably because
  357. 10:18it's such a new topic. But, I'm going to
  358. 10:20put on screen now a couple videos that
  359. 10:22I've done and some other resources as
  360. 10:23well. If you guys want, I can also make
  361. 10:25a video that specifically covers how to
  362. 10:27do agentic engineering. You know what's
  363. 10:29crazy is that to learn all of these
  364. 10:31topics that we just discussed prior to
  365. 10:33AI, it would have taken you years. go to
  366. 10:35school for this and then work multiple
  367. 10:37years to get to this level. Or, even if
  368. 10:38they're attending like a fast-track boot
  369. 10:40camp, we're still talking like at least
  370. 10:4210 to 12 months completely full-time.
  371. 10:44But, with good learning techniques and
  372. 10:45with AI to help with your learning, you
  373. 10:47can reduce the learning time down by so
  374. 10:49much. I think that if you really focus
  375. 10:51and if say you're doing it like
  376. 10:52completely full-time, I would estimate
  377. 10:54maybe like 6 months to be able to learn
  378. 10:56all these topics very solidly. So, in
  379. 10:58the next section, I'm actually going to
  380. 10:59share some of my favorite learning
  381. 11:00techniques that can help you learn
  382. 11:01faster. For those of you who have
  383. 11:03watched my channel for a while, you
  384. 11:04probably also know that learning things
  385. 11:06fast is like my favorite activity. So,
  386. 11:08very exciting. Let's now talk about tips
  387. 11:10to supercharge your learning.
  388. 11:12I'm going to divide these tips up into
  389. 11:13two different categories. The first is
  390. 11:15how to use your resources better, like
  391. 11:16the lectures, the books. And the second
  392. 11:18is how to use projects for learning and
  393. 11:20specifically how to use a AI coding
  394. 11:21agent to help you build a project and
  395. 11:24learn at the same time. Two birds with
  396. 11:26one stone. Okay, let's first talk about
  397. 11:27how to use resources. If your resource
  398. 11:29happens to be like a video or like some
  399. 11:31text or like readies and stuff like
  400. 11:33that, I would really recommend using
  401. 11:35something like Notebook LM and put all
  402. 11:37that information in there and ask it to
  403. 11:39summarize things for you and also to
  404. 11:41generate questions that will guide you
  405. 11:43along the work. This is called primate.
  406. 11:45It can really speed up your learning
  407. 11:46process because you understand what are
  408. 11:49the things that you need to be learning
  409. 11:50and how they fit together with each
  410. 11:51other. As you're going through the
  411. 11:52lectures, the text, and like code
  412. 11:54snippets and and whatever, um if there's
  413. 11:56anything you don't understand, you can
  414. 11:57also just copy paste that and put that
  415. 12:00into AI to help you explain what's going
  416. 12:02on. When it comes to code and logic, at
  417. 12:04least of the filming right now, the best
  418. 12:06AI model to do this would be Claude. I
  419. 12:08personally default using Claude's on it
  420. 12:09for anything that happens to do with
  421. 12:11code. But if you don't want to pay for
  422. 12:12it and you run out of the free trial,
  423. 12:13honestly most modern AI models would be
  424. 12:16able to help explain code. You can even
  425. 12:18use the free ones, the open source ones,
  426. 12:19like Gwen, Kimmy, or Deep Seek. Another
  427. 12:21way that I really like using AI models
  428. 12:23to help me learn is by taking a concept
  429. 12:25and ask it to give me examples and
  430. 12:26analogies because this will really allow
  431. 12:28it to solidify my mind to understand the
  432. 12:30concept. Like for example, if you're
  433. 12:32learning about object-oriented
  434. 12:33programming, you can ask it explain
  435. 12:35object-oriented programming to me using
  436. 12:36analogy and provide examples. Much
  437. 12:38clearer, I think. And finally, often
  438. 12:40times the best way to learn how to code
  439. 12:42is to take the code and the snippets
  440. 12:44from the resources that you have, like
  441. 12:46other people's code, and then ask AI to
  442. 12:48help you understand the code that's
  443. 12:50there, to explain to you what the code
  444. 12:52is doing. This is so powerful and I'm
  445. 12:55kind of salty that I did not have this
  446. 12:58option when I was learning how to code.
  447. 13:01To do this, you can also just like copy
  448. 13:02paste or like upload the code into an AI
  449. 13:04model, or this could be a good
  450. 13:05opportunity for you to set up an AI
  451. 13:07coding agent. Just give it the code and
  452. 13:09it would explain to you the structure of
  453. 13:10the code, how the pieces are fitting
  454. 13:11together, um what different lines of
  455. 13:13code mean, line by line if you want. And
  456. 13:15speaking of projects, this is a good
  457. 13:16transition to the second category of
  458. 13:19tips that I have, which is for projects.
  459. 13:21First of all, I want to just emphasize
  460. 13:22that projects are the best way for you
  461. 13:25to learn how to code, hands down. You
  462. 13:26can watch like so many lectures, read so
  463. 13:29many things and whatever, and you think
  464. 13:30that you know these things. Yeah, but
  465. 13:31you don't unless you actually start
  466. 13:33building with it. And the way you do
  467. 13:34that is by doing projects. So, I
  468. 13:36encourage you to do projects
  469. 13:37continuously as you're learning. Like if
  470. 13:39you're learning about how to use an API,
  471. 13:40actually make a project in which you use
  472. 13:42an API or multiple APIs. If you're
  473. 13:44learning about testing, write some
  474. 13:46tests. Try to understand the differences
  475. 13:47between different tech stacks. Try them
  476. 13:49out yourself. Especially with the help
  477. 13:51of AI coding agents, it is so easy to do
  478. 13:53this. You can ask the coding agent
  479. 13:55something like build me a music storage
  480. 13:57software with like XYZ specs built with
  481. 14:00a Django back-end and Next.js front-end
  482. 14:03and store the data in a SQLite database,
  483. 14:05for example. This is how you're going to
  484. 14:06learn faster, gain more experience, and
  485. 14:07also work towards using AI coding agents
  486. 14:10with agentic engineering. I previously
  487. 14:12mentioned other people's projects as
  488. 14:13well. Another great way for you to learn
  489. 14:15is to take other people's projects and
  490. 14:18then adopt them. Add on more stuff to
  491. 14:20them. Change around the code and see how
  492. 14:22the results change. This is such a great
  493. 14:24way to learn and it's people amongst
  494. 14:25engineering students everywhere. Another
  495. 14:27tip, as you're coding, you can also ask
  496. 14:29AI to give you multiple suggestions for
  497. 14:31how to implement something. This way you
  498. 14:33can expand the scope of your knowledge,
  499. 14:35which is going to make you a much better
  500. 14:36coder. And final little tip, ask AI to
  501. 14:38help you understand documentation. I
  502. 14:40hate reading documentation. Like I hate
  503. 14:43it so much. I don't think I've read any
  504. 14:45documentation since like 2024 cuz I will
  505. 14:48ask AI to help me understand
  506. 14:49documentation. To be clear, I still need
  507. 14:52to understand the documentation. I just
  508. 14:53don't like reading it. So, this is a
  509. 14:55great way to understand documentation
  510. 14:56faster and incorporate them into your
  511. 14:58projects. I'm not going to go into Vibe
  512. 15:00Coding side agentic engineering specific
  513. 15:02tips right now cuz there's just like way
  514. 15:04too much and there's a whole other video
  515. 15:06that I can make on this topic. So, I'm
  516. 15:08just going to put on screen some of my
  517. 15:10favorite little Vibe Coding tips. If
  518. 15:13that's helpful for you guys, just know
  519. 15:14that this is not everything and it might
  520. 15:16be confusing if you never had any
  521. 15:17experience at all. Let me know in the
  522. 15:19comments if if want me to actually make
  523. 15:21like a fully fledged video uh for that
  524. 15:23topic. All right, we have now come to
  525. 15:25the end of this video. Thank you so much
  526. 15:27for watching until the end. I know that
  527. 15:29was a lot. I really hope this is helpful
  528. 15:31to you. I'm so glad that you're learning
  529. 15:33how to code in 2026. Best of luck in
  530. 15:35your coding journey and I will see you
  531. 15:36guys in the next video or live stream.

About this transcript

This page contains the full transcript of How To Learn To Code In 2026 by Tina Huang, generated from the public captions YouTube serves with the video. The transcript has 3,655 words across 531 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.