# Grok 4.7 + Grok Bot Is CHEATING (Only $20 NOW!)

Channel: Parker Prompts
Video: https://www.youtube.com/watch?v=OPThXcTW9Rk
Duration: 11 min
Language: English
Words: 1991
Transcript page: https://viewrankai.com/tools/youtube-transcript/OPThXcTW9Rk

---

[0:00] If you're still using chat GPT or Claude and have no idea what Grok can do now, you're already behind because while you're copying answers out of a chat window, people on Grok Bot are waking up to finished work they never touched. So, in this video, I'll show you exactly what you can build with Grok and how to get all of it for just $20 a month. And the reason Grok is worth a second look right now is Grok 4.7 [snorts] which xAI released not that long ago. It was trained on hard tasks that take hours to finish and it checks its own work before handing anything back. It

[0:40] also holds up to 500,000 tokens at once, so it can read through huge documents or whole code base without losing track. In xAI's own benchmarks, it scores just behind Claude 5.1 at multi-hour office work and it's ahead of GPT 5.6 Soul in that regard. It also beats GPT 5.6 Soul at coding, although Fable still leads there and wins four of the seven benchmarks xAI published. But the price is where they are far apart because through the API, Fable costs around eight times more than Grok 4.7 for the same output. On top of that, it was trained specifically to work inside Grok Bot, which is what makes the rest of

[1:23] this video possible. That's an app where you create AI assistants called bots and you message each one the way you text someone on your team. The big difference between Grok Bot and the traditional LLMs we see is that it's more like an AI agent that has its own computer, so it can see and click through stuff just like you [music] would and sign onto your tools right in the browser. And it keeps working even after you close your laptop. The other piece of this whole thing is Cursor, which is an AI coding tool and its $20 pro plan is what gets you both Grok Bot and Grok 4.7. But none

[1:58] of this matters until you see how it handles real-world tasks. And before we jump in, if you want to go from casually using ChatGPT to actually building systems like the ones I show in my videos, that's exactly what I teach inside AI Fluency. It's my 12-week program and school with a new video every day and over thousands ready-made resources, and the link to join is in the description. The first use case I

[2:23] tested Grok on is finding new clients. Because if you freelance or run an agency, that's the part of the job that never ends. To find a good lead, you have to go through brand after brand, open each website, work out if they'd actually be a good fit, and then write a pitch that doesn't read like every other cold email. And if you want it to stand out, you'd show them a quick mock-up of what their site could look like, which takes even longer. I wanted that whole process to run without me. So, I'll create a new bot and paste in this prompt. The prompt sets the bot up for

[2:57] Field Note Studio, an agency I made up for this video, and describes our style. Then it asks the bot to find three real brands that could become our clients, meaning they have great products, but a generic website. For each brand, it redesigns their homepage in our style through a Cursor Cloud agent on Grok 4.7, publishes it to a live link, and drafts a pitch in Gmail with that link without sending anything. Once it's running, the bot starts going through websites on its own computer. It found Field and Forge, Sagebrush Dry Gear, and Week of Hope, and the redesign for each one is already live. Keep in mind, these

[3:36] are real companies, and what you're seeing is the new homepage with pitch to them, not their actual website. So, let's go through them, starting with Field and Forge. This one is really well-built with clear benefits, all the details, and sections that explain exactly what they sell. Seg brush dry gear follows a similar layout and it looks just as clean. And the week of hope is a little different because this time the page focuses on the [music]

[4:03] product itself instead of its benefits. On top of that it checked every page on desktop and phone, locked them in a Google sheet and saved the pitches as drafts in my Gmail. From there I can open [music] any of them, see what it wrote and edit or send it as it is. And that's what makes this so useful. Finding the right brands and building something to show them is the slow part and it's already done. So all I have to do is read and hit send. Then I'll set it to run twice a week so new leads keep coming in while I sleep. So from one

[4:35] single message I've got three real leads with mockups built for them. Plus the emails already written and I didn't have to touch any of it. And the next use case is for anyone running a business in a crowded market who wants to keep track of their competitors pricing. To do that I'll create a new bot and paste in this prompt. It's set up for Echo, [music] an AI note keeper startup I made up going up against seven real competitors like Otter, Fireflies and Granola. The bot checks all their pricing pages and change logs, logs everything in a Google sheet and has wrote 4.7 built a live

[5:14] dashboard out of it. The cloud agent had it ready in about 18 minutes. So if we open it up we get this really beautiful page with all the data already loaded in. First there's a timeline of what changed like Fireflies launching free unlimited voice dictation in its desktop app on September 29th. And right under it are two more recent updates from Granola and Avoma. Below that is a chart what five seats would cost you per month with each of them. Granola is the cheapest at $70, Avoma and TLDR both cost the most at $145, and the rest are mostly between $85 and $100. Then, each competitor gets its own

[5:55] card showing every plan's monthly and yearly price per seat, plus what the free plan actually gives you. And the last part is a feature grid that lines up the top three features of all their plans side by side. So, instead of checking seven websites a day, I've got one dashboard that stays up to date without me. [music] And the next bot fixes a problem almost everyone has on YouTube. Whenever you come across a video you want to watch and learn from, you save it to your watch later list because you don't have the time right now. But over the time, you never really go back to it because there are just so

[6:31] many videos in there, and you don't know which ones are actually worth watching. So, to fix that, I'll create a new bot, take over its computer, and sign onto my YouTube account through its browser. Once signed in, I'll paste in this prompt. It tells the bot to go through my watch later, read the full transcript of each video, and write up a summary, the key timestamps, a short quiz, and a verdict on whether to watch them full. And after all that is written up, Grok 4.7 turns it into a learning site. So, when it's finished, I'll open up the link it sends me. That takes me to this beautiful

[7:07] dashboard with all my videos on it, where I can search them or just click through the cards. Each card also has a label telling me whether to watch it or just skim it. It starts with a quick summary of what the video is about, followed by the key steps with timestamps and to one specific action to

[7:24] take. [music] And at the end, there's a flashcard quiz to test how much I remember. Plus, if I want to see the full video, the watch on YouTube button right up top takes me straight to it. This way, I can go through my huge pile of watch later videos quickly and cut my learning time in half. Up until now, every bot had one fixed job, but this next one can make pretty much anything you ask for. So, I'll create one more bot and paste in this prompt. It sets the bot up as a tech lead that plans the work and hands

[7:57] the coding to cloud agents on Grok 4.7. Before building anything, it has to tell me what it thinks I want and ask me up to three questions. That stops it from guessing and building the wrong thing. Then, it splits the build into small tasks and gives each one to its own cloud agent. When they're done, it tests the live page itself and sends anything broken back to be fixed. And it only

[8:21] messages me when it's stuck or finished. For my first request, I asked for a focus timer that makes me want to use it. It's a simple Pomodoro timer where I pick how long to focus and hit start. And as the time goes by, the scene grows and the sun moves from morning to evening. The whole thing is minimal and genuinely something I'd use. You can ask for a landing page or a small business tool the exact same way. So, now I've got a bot I can just tell what I want and it figures out what needs to be done and builds it on its own. And as you

[8:54] keep making bots like these, they start to pile up and it gets hard to keep track of what each one is doing and what each one is meant for. So, to fix that, I made a chief of staff that manages all the other bots for me. Because it knows which bot owns which job, I can ask it for anything and it will send the work to the right [music] one. And if something needs more than one of them, it splits the request up, passes the results along, and gives me back one short answer. It never does any of the job itself. It just keeps the whole team

[9:25] moving and sends me a summary every morning of everything they got done. This works because bots in Grok Bot can message each other directly, so the chief of staff talks to them the same way I would. And since they all run on the same cloud computer, the the links and files one bot makes [music] are right there for the others. To put it to the test, I asked it to get a status update from Client Finder, Market Watch, and Watch Later, and then have Builder turn all of that into one page. Without me stepping in once, it messages all three, collects the replies, and handed the brief over to Builder. It's pretty

[10:01] clean-looking dashboard with just three cards and I can see at a glance exactly what each bot is doing. Now, to be fair, this HQ page isn't something you really need, but it does show that from a single chat, you can get every bot to reporting, hand to work to each other, and build something new out of it. And if you want help building a team like this for your own work, that's what the

[10:24] community inside AI Fluency is for. Right now, it's $96 a month, and the price goes up every 50 to 100 members. So, join AI Fluency with the link in the description and keep today's price before it goes up.

---

About this transcript
Read from YouTube's own caption track and laid out by ViewRank AI (https://viewrankai.com).
ViewRank AI finds the videos already beating a creator's own average on Instagram, TikTok and YouTube Shorts, transcribes them from the audio itself in more than 60 languages, and turns what worked into new ideas and scripts.
Free transcript tools, no account needed: https://viewrankai.com/tools
How to read any video this way: https://viewrankai.com/llms.txt