Loading video...

Video Failed to Load

Go Home

Claude 3.7 Sonnet is the best model in the world for code. It’s like having a world-class dev with exceptional taste. “build a next.js SaaS marketing template” And boom - 26 files of beautiful code in one shot. Absolutely unreal.

547,157 views • 1 year ago •via X (Twitter)

10 Comments

Mckay Wrigley's profile picture
Mckay Wrigley1 year ago

Skip to 1min and just watch how good the template it generated is… THAT’S IN 1-SHOT FROM A 1-LINE PROMPT Taste is the next frontier for these models.

Mckay Wrigley's profile picture
Mckay Wrigley1 year ago

Also - Claude Code CLI is awesome! Have had a chance to test it for a bit and it’s obvious to me that more labs should be doing CLI tools (agents?). Would love to see every lab make one. The possibilities there are pretty profound. Labs that are dev 1st will default do well.

Mckay Wrigley's profile picture
Mckay Wrigley1 year ago

This release should *really* make you feel the AGI. Throw a couple of these guys into agentic coding systems and you have an autonomous product system. If you haven’t been taking coding agents seriously you should begin doing so. It’s 100% here.

Christian Esmann's profile picture
Christian Esmann1 year ago

I just launched an all-in-one template for cross-platform development, based on the stack I use myself everyday. Expo, NextJS, TypeScript, Tailwind, Firebase, AppsFlyer, Authentication, Analytics, In-App Purchases, Stripe and a lot more, setup by default.

Alex Albert's profile picture
Alex Albert1 year ago

yeah the UI generation ability of this model has really blown me away

Mckay Wrigley's profile picture
Mckay Wrigley1 year ago

Taste too cheap to meter :)

dennis's profile picture
dennis1 year ago

Then we have to ask that question: Claude 3.7 or o1 pro for coding?

jj_investor 🍓's profile picture
jj_investor 🍓1 year ago

Haven’t got access yet but Grok3 is off the hook

Jamie Munro's profile picture
Jamie Munro1 year ago

Anthropic said they "optimized somewhat less for math and computer science competition problems, and instead shifted focus towards real-world tasks that better reflect the needs of our users." So it will be interesting to see how it performs for tasks that aren't coding.

compuficial's profile picture
compuficial1 year ago

we really moved on quickly from Grok 3

Related Videos