Opus 4.6 vs Gemini 3.1 Pro (2/9) prompt: Build a premium Vite + React landing page for Aether, an AI creative studio, with dark-mode design, high-end animations (Framer Motion), and conversion-focused sections (hero, features, demo, pricing, testimonials). Whilst I think gemini is overall better at design, the model often... struggles with syntax issues, tool calling and thinking for 2-3 times the amount of opus.show more

Gemini: "Here's your landing page." ✅ Opus: "Here's your landing page, your brand identity, and your next 3 campaigns." 😂

Gemini: prettier mockups Opus: actually runs without 47 syntax errors Pick your fighter

I loved this one

landing pages are kind of the easy case for both models. would be more interesting to see how they handle complex state management or form validation edge cases

Opus 4.6 landing looks like AI slop Gemini worked way better imo

Gemini looks less AI

Same purply shit

What harness did you use for the Gemini model?

gemini looks slick but opus just ships faster tbh

@grok 你搜集下最近的帖子，看下gemini 3.1pro 好，还是 opus 4.6好，

Thank you for sharing, this is great insight 🤝

i think gemini is better

Objectively speaking, they are the same. But visually, I like Gemini more.

give them the front end skills and redo this test PLS

You know what I think of the vibe code apps? How do you actually get them working.

i’m finding the output of Claude nicer than Gemini. but Claude doesn’t do any image gen

I’m using Gemini — a professional full-stack model, and it costs about 3× less than Opus.👌👌

4.6's incremental dev tools often outmatch raw outputs

opus runs, gemini glitters

they both nailed the smooth tab transition and Gemini 3.1 is a bit better.

I don't understand these gadrient AI slops both models are displaying

Gemini 3.1 context is elite, but Opus 4.6 still holds the edge on reasoning for complex Framer logic.

how about this one ?

@Dont_Be__Lazy Focused

Check this: I made a few nextjs templates!

Got lost in the nuances of that comparison. So much potential.

people keep arguing benchmarks but this is what actually matters: one model designs prettier, the other actually ships working code

Looks like opus is scrolling faster than Gemini 😇

Do you think there could be a case for a hybrid workflow here — using Gemini for the design/planning phase (layout, structure, visual decisions) and then handing it off to Opus to actually execute and write the code? Best of both worlds?

oh man, i've been testing both for travel planning lately. opus handles complex itineraries better, but gemini is faster for quick edits. depends if you need deep reasoning or speed.