Navigate Select ESC Close

I Let 3 AIs Compete to Build the Same App…

2025-11-01 Education
4.3m
1.0k
36
Tech With Tim
Tech With Tim
2.0m subscribers

Unlock all features

FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.

Description

Today, I'm putting three of the most hyped AI coding platforms to the test: Blitzy, Devin, and Factory AI. Now, each of these tools take a slightly different approach to building software. I'm going to give all three of them the exact same real world task, something the developers actually do in companies all the time. We're going to see which platform performs the best, and how much effort it takes from my end, the programmer. Want to make real money with coding? I share high-signal insights on careers, monetization, and leverage in my free newsletter. Join here and get my guide How to Make Money With Coding instantly: https://techwithtim.net/newsletter 🎞 Video Resources 🎞 Blitzy: https://blitzy.com/?utm_source=YouTube&utm_medium=Social&utm_campaign=Tech%20with%20Tim Factory: https://factory.ai/ Devin: https://devin.ai/ AWS Card Demo Repo: https://aws.amazon.com/blogs/opensource/introducing-open-source-aws-carddemo-for-mainframe-modernization/ Blitzy SWE Bench: https://paper.blitzy.com/blitzy_system_2_ai_platform_topping_swe_bench_verified.pdf Devin SWE Bench: https://cognition.ai/blog/swe-bench-technical-report Factory SWE Bench: https://factory.ai/news/code-droid-technical-report ⏳ Timestamps ⏳ 00:00 | Overview 00:52 | Demo Task 02:31 | Prompt 03:20 | SWE Bench Results 05:50 | Blitzy 11:35 | Devin 15:58 | Factory Hashtags #BlitzyAI #CognitionDevin #FactoryAI

Top Comments (10)

@AkhilJaini 2025-11-02

I don't understand this comparison. Are you just assuming Blitzy is better than the others because it has infinite context, takes longer and only requires 1 prompt? What should actually matter is the final result, whether the new code it generated actually works as expected or not, none of which you show in this comparison. Otherwise, what even is the point of these AIs...

19 1 replies
@TimMountjoy-zy2fd 2025-11-02

We need the results to be fully tested and report on which does the best job once finished. We are assuming Blitzy is the best cos it finished everything in one hit but without testing we are guessing.

10
@nkofeebit 2025-11-01

can you compare to cursor? or is it incomparable as cursor is using generic LLMs but those planform most likely some dedicated?

4 1 replies
@IsniTech 2025-11-01

The best coding teahcer, you are helping me start my journey!

1
@nyfromla 2025-11-17

I have investigated and spoken with blitzy, and it is pretty impressive (or at least the promise of what it does). I hoped, at the end of the video you actually showed us the end result for each. I know you don't have three build teams running each solution, so the final deliverable for each and time to finished code isn't realistic. but I have yet to see the UI. here is what it ended with. Since UI and UX are critical to my teams, I'd love to see how it interpreted the UX. Is this a planned video in the future?

0
@CalConrad 2025-11-09

Please try Blitzy with a big greenfield (no codebase) project (like a SaaS) and actually run the code to see how it does.

2
@sachetagarwal6141 2025-11-06

You really forgot to do basic on Devin which is create DeepWiki like what you did for Blitzy.. was it intentional??

0
@emmanuelalem7448 2025-11-03

very helpful content💯💯

1
@AuditingSC 2025-11-18

Think it funny they changed the name of these programs from "Generator" to "AI" and everyone falls for it. Shouldn't surprise me really considering everyone feel for the "Smart" epidemic Smart Phone, Smart Cars and my Favorite Smart Water.. lol

0
@HumaylImran 2025-11-01

Hey TechWithTim! Can you please make a 2025 python full course tutorial?

5

Unlock the Data Inside
Turn Videos into Knowledge

  • Get FREE 10/day: transcripts, summaries, chats
  • Chat with videos, export text & PDF
  • $1 free API credit for RAG, chatbots & research

Free forever plan • All features unlocked

App screenshot