So Grok 4.5 just released on Cursor

@omarontape

88.2K followers · usually 16.4K views a post

Views

44.8M

Likes

68.6K

Comments

152

Shares

189

Selling · SaaS

Cursorseen in 62 videos →https://cursor.com/ ↗

Why it worked

The video uses a high-stakes, real-world problem (a critical app bug) to provide a transparent, data-backed comparison of two AI tools, making the promotion feel like a helpful technical case study rather than a traditional ad.

Summary

The creator tests the new Grok 4.5 model within the Cursor IDE to fix a critical bug in his own app, Vlogit. He compares the performance and cost of Grok 4.5 against Opus 4.8, demonstrating that Grok is significantly faster and cheaper for both coding and documentation tasks.

Structure

1Introduce the new Grok 4.5 model on Cursor
2Explain the critical bug in the Vlogit app
3Compare Grok 4.5 and Opus 4.8 performance on the bug
4Compare the cost and efficiency of both models
5Showcase non-coding task performance
6Provide a call to action for the product

Original caption

Follow me then comment “Cursor” to receive 50% off your first month of Cursor, double grok 4.5 usage for a week, and a link to download Vlogit! Since Grok 4.5 just came out and it’s supposedly way cheaper and faster than Opus 4.8, I decided to test it on a really complex bug I had in Vlogit. In our system, when people post vlogs, we queue them for processing and transcoding to ensure a better and faster playback experience for our users. But the other day when I checked the queue, I saw 30k waiting when it’s typically in the tens, so something went heinously wrong. I had both Grok 4.5 and Opus 4.8 investigate and fix the issue. While both reached the same result, Grok ended up being 5x cheaper and 2-3x faster, using less than half the tokens. #CursorCreator

More from @omarontape

Their other posts in the index, biggest breakout first.

Product placement

The video features the Cursor IDE interface throughout the demonstration of the bug fix and documentation creation.

Call to action

Comment "Cursor" to receive 50% off your first month of Cursor, double Grok 4.5 usage for a week, and a link to download Vlogit.

On-screen text

So Grok 4.5 just released on Cursor and apparently it's faster than Opus 4.8 Build a weather app 31,232 vlogs in queue CURSOR COMMENT "CURSOR"

Transcript

So Grok 4.5 just released on Cursor and apparently it's much cheaper and faster than Opus 4.8. So what better way to put it to the test than using it to fix a bug I had in my app, Vlogit. I launched the app 3 weeks ago and we're already at 40k users, but the other day I noticed a huge bug: 30k vlogs were waiting to be processed. Typically the queue reaches like 100 max. So it turns out one corrupted file was freezing the queue for every single vlog uploaded after it, causing like a traffic jam and it caused playback to be much slower across the entire app. So, I gave this exact same bug to both models in Cursor. Grok diagnosed the issue and fixed it in one shot. It took about three and a half minutes and about 100k tokens. And Opus got there too, but it took 20 minutes and burned 250k tokens. And since I'm always looking for the most efficient options, I decided to check the cost too. Grok costed about 65 cents while Opus costed more than $5 dollars... 8x the price for the same result. I also wanted to see how they fared in non-coding tasks too, so I had them both create detailed documentation for the fix so that it wouldn't happen again in the codebase. And again, Grok finished in a minute using 12k tokens while Opus took 3 minutes using more than 20k tokens. It's a completely different weight class than even Composer 2.5, just looking at the benchmarks. Vlogit's gonna be bigger than BeReal, so follow along for the journey. And comment "Cursor" and I'll send you 50% off your first month, double Grok 4.5 usage for a week, and a link to download Vlogit.