YouTube2Text

I fixed Claude Code Usage Limits. Here's how... — Transcript

by Albert Olgaard · 1,542 words · 206 segments · language en · Watch on YouTube

Full transcript

  1. 0:00Cloud Code is really expensive. Even on
  2. 0:03my $200 a month plan, I'm still running
  3. 0:06into rate limits. If you don't know me,
  4. 0:07my name is Albert Olcott. I didn't go to
  5. 0:10university, but still managed to build
  6. 0:12two million-dollar companies only
  7. 0:14because I know how to use AI. Now I'm
  8. 0:16making AI education accessible for
  9. 0:18absolutely everyone completely for free,
  10. 0:20and I run one of the largest AI
  11. 0:21communities in the world with 150,000
  12. 0:24members all learning AI. What I'm about
  13. 0:26to show you will save you a lot of
  14. 0:27money. How can you use Cloud Code for
  15. 0:29free? Well, Cloud Code is what's called
  16. 0:31an agent harness, which means that Cloud
  17. 0:33Code is really just a wrapper. The real
  18. 0:35brain is the actual AI model. So when
  19. 0:38you make a response, this goes to Cloud
  20. 0:40Code that wraps it, and then it sends it
  21. 0:42to the AI model. And if the AI model is
  22. 0:44the brain, then the harness is kind of
  23. 0:46like the body that can call your tools,
  24. 0:49run loops, manage memory, handle errors,
  25. 0:51and then it's also in the harness that
  26. 0:52can actually read and write the files.
  27. 0:54So the AI model, the brain, is what
  28. 0:56controls, let's say, the arm, which is
  29. 0:58what actually changes your code. When
  30. 1:00you're paying for Cloud Code, you're not
  31. 1:01paying for the wrapper. The only thing
  32. 1:03that you're paying for is the actual
  33. 1:05brain. It's the model, which means if we
  34. 1:06connect to another model that's cheaper
  35. 1:09or even free, we can actually use Cloud
  36. 1:11Code completely for free. I've created
  37. 1:13this folder that I called free Cloud
  38. 1:14Code, and I'll leave a link to this
  39. 1:16Google Drive right below this video.
  40. 1:18Then let's open Visual Studio Code,
  41. 1:20click open, and choose the folder that
  42. 1:21we just downloaded from the drive, click
  43. 1:23open, and click yes, I trust the
  44. 1:25authors. And basically there's nothing
  45. 1:26inside of this folder other than the
  46. 1:28actual Cloud Code structures. You don't
  47. 1:30need to change anything yourself.
  48. 1:31Usually when you spin up Cloud Code and
  49. 1:33you write Cloud, you'll be able to see
  50. 1:34what model it's using. Right now it's
  51. 1:36using Opus 4.6, and I'm on my Cloud Max
  52. 1:39plan. So you can see the model and how
  53. 1:41it charges you. If I write {slash}
  54. 1:43model, the only models that I can choose
  55. 1:45from is Haiku, Sonnet, and then Opus,
  56. 1:48where Sonnet and Opus are both pretty
  57. 1:50expensive models compared to what else
  58. 1:52is on the market. So you're very capped
  59. 1:53with just a regular Cloud Code
  60. 1:55subscription. Instead, if I go back into
  61. 1:57this folder and I write Cloud, you can
  62. 1:59see we're using a different model. We're
  63. 2:01using the new Google model, the Google
  64. 2:03Gemma 4, which is an open-source model,
  65. 2:05and you can see it says API usage
  66. 2:07billing instead. I've basically managed
  67. 2:09to change the model inside of Cloud Code
  68. 2:11to any model that I want. Let me show
  69. 2:13you how to do that. I just want to make
  70. 2:14one thing clear. If I write {slash} log
  71. 2:16out, it's going to log me out of Cloud
  72. 2:18Code, and I then write Cloud again.
  73. 2:20There's two real ways to get access to
  74. 2:22Cloud Code. Either you need a Cloud
  75. 2:23account subscription, or you can use the
  76. 2:25API billing. If you really want to save
  77. 2:27a lot of money, then you can get access
  78. 2:29to Cloud Code by choosing the console,
  79. 2:30adding $5, and you can run it completely
  80. 2:32for free forever. The reason that we
  81. 2:34downloaded this driver is because if I
  82. 2:35click down on the {dot} Cloud folder,
  83. 2:38and then hit inside of settings {dot}
  84. 2:40local {dot}
  85. 2:44right here that we have set, which is
  86. 2:45not normal to do inside of Cloud Code.
  87. 2:47And inside of here we have changed the
  88. 2:49URL to make API requests, and we've also
  89. 2:51added a different auth token. In order
  90. 2:54to do this, you need to remove this
  91. 2:55token that's already in here, and don't
  92. 2:56worry, I've already rotated the API key.
  93. 2:58Then you need to go inside of
  94. 3:00OpenRouter, which is basically a site
  95. 3:02where we have access to all the models
  96. 3:03that exist. Go inside of settings, then
  97. 3:06go to API keys, create a new API key
  98. 3:08that we just call Cloud Code, copy this
  99. 3:10API key, go back into this settings
  100. 3:12{dot} local {dot} JSON file, and paste
  101. 3:15this in as the entropy auth token. And
  102. 3:18now we can change whatever model we want
  103. 3:19to use. By default, it's going to have
  104. 3:21the Gemma 4 model, which is a really
  105. 3:23good new open-source model that Google
  106. 3:25just dropped. So if I save this file and
  107. 3:28click control save, every time you do
  108. 3:30this, you need to restart Cloud Code, so
  109. 3:31click control C, write clear, write
  110. 3:34Cloud, and then I write, "Hi, can you
  111. 3:37hear me?" It's now going to be using the
  112. 3:39Google Gemma 4 model. Let's say that you
  113. 3:41are on the $20 a month plan right now,
  114. 3:43and you are running low on usage. For
  115. 3:45easier tasks, you can just change the
  116. 3:46model, and you can see the Gemma model
  117. 3:48now says, "Yes, I can hear you. How can
  118. 3:50I help you today?" That is this model
  119. 3:51right here, which is extremely cheap. We
  120. 3:53can also search for GPT 5.4, that is
  121. 3:57Open AI's model, and all you do is that
  122. 3:59you just copy the model name right here.
  123. 4:01Then we go back into the settings {dot}
  124. 4:02local {dot} JSON, and then we change all
  125. 4:05of these models. Hit control S to save
  126. 4:08it, and now we have to restart Cloud
  127. 4:10Code. And now we can see it says GPT
  128. 4:125.4. If I write, "Hi, can you hear me?"
  129. 4:15it's now going to be using the GPT 5.4
  130. 4:17model, and it says, "Yes." Now you know
  131. 4:19how to change any model you want inside
  132. 4:21of Cloud Code. But let me show you
  133. 4:22something that is really, really cool,
  134. 4:24cuz these models that I've shown right
  135. 4:25now still does cost something. If you go
  136. 4:27inside of OpenRouter, you click explore
  137. 4:29models, it's going to show you all of
  138. 4:30the models that are inside of here. If
  139. 4:32you go to prompt pricing, and you just
  140. 4:34set it to free, it's going to show you
  141. 4:35all of the free models that are right
  142. 4:37now inside of OpenRouter. For example,
  143. 4:39this Qwen 3.6 Plus is one of the best
  144. 4:42models out there right now, and you can
  145. 4:43test if it is working by going to the
  146. 4:45chat and writing, "Hi." See, it says,
  147. 4:47"Hi there, how can I help you today?" So
  148. 4:48this model is working. What I'll do is
  149. 4:50that I'll go back, and then I'm going to
  150. 4:52copy the model name, go back into our
  151. 4:55settings {dot} local {dot} JSON, and
  152. 4:56then change this model. [music] And you
  153. 4:59can see now it's using the Qwen 3.6 Plus
  154. 5:02free model, and you can see it says,
  155. 5:03"Yes, I can hear you. How can I help?"
  156. 5:05And I don't have to pay anything for
  157. 5:06using this model. If we go inside of our
  158. 5:08OpenRouter and inside of logs, we can
  159. 5:10see how much each request actually cost
  160. 5:13us. You can see the Gemma 4 model was
  161. 5:15costing us a little bit, the GPT 5.4
  162. 5:17models were costing us a little bit, but
  163. 5:18this Qwen 3.6 Plus is completely free.
  164. 5:22Is it as good as the paid models? No, of
  165. 5:24course not. You always get what you pay
  166. 5:25for, but this is a really good
  167. 5:27alternative if you hit a rate limit and
  168. 5:28now you need to use something else. But
  169. 5:30what do you actually do with this
  170. 5:32information? When you hit a rate limit
  171. 5:33all the time, people on Twitter are
  172. 5:35going crazy about hitting rate limits
  173. 5:37all the time. just switch to a cheaper
  174. 5:39model that is still going to perform
  175. 5:40very, very well. Like if you hit a rate
  176. 5:42limit on Cloud Code, you can switch to
  177. 5:44something that is very cheap, like Gemma
  178. 5:464, or something that is a little
  179. 5:47cheaper, I think like 50% of the price,
  180. 5:49like GPT 5.4. If you want to learn more
  181. 5:52about Cloud Code and more about building
  182. 5:53an AI agency in general, where you are
  183. 5:55selling this as a service to businesses,
  184. 5:57then you can check out our community,
  185. 5:59The 1% in AI. Inside of here we have
  186. 6:01every single thing that you need to
  187. 6:02learn in order to start an AI agency,
  188. 6:04and we are soon dropping a Cloud Code
  189. 6:05masterclass, which is hours of content
  190. 6:07specifically on Cloud Code. And the best
  191. 6:09thing is that right now we are running
  192. 6:11the 30-day challenge that gets you from
  193. 6:13nothing to an AI agency in 30 days. And
  194. 6:15the best thing is when you stay
  195. 6:16consistent and you finish the 30-day
  196. 6:18challenge, we actually get your first
  197. 6:19month completely refunded, which means
  198. 6:21if you are committed, you get to join
  199. 6:23completely for free. If you want to
  200. 6:25learn more about how that works, then
  201. 6:26I'll leave a link right below this
  202. 6:27video. Thank you guys so much for
  203. 6:29watching. You two things that this video
  204. 6:30is going to be interesting for you. As
  205. 6:32always, have an awesome rest of your
  206. 6:33day. Albert out.

About this transcript

This page contains the full transcript of I fixed Claude Code Usage Limits. Here's how... by Albert Olgaard, generated from the public captions YouTube serves with the video. The transcript has 1,542 words across 206 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.