Chapi Dev Talk Group.

Sol on high effort is amazing during planning and review sessions. It asks really good grilling questions and can reason well when making decisions. It also does great for high-level product specs and general plans. I still reach for fable when I want to write up implementation plans though. My implementers are either Sonnet 5 or Terra with not too much of a difference in quality or token spend

Incredible!

Why a fronteir model as an executioner?

So about the token usage can you eski check in claude web, Ask it to create the claude mds with critically economical token usage😁 It really did help idk hw but it did

Since its cheaper and is really great at going all the way to do a lot of debugging

Yes Sir 😂

If you have some private abandoned projects, submit them here and make some money from them at least: https://atrium.afterquery.com/r/CSVN3P

♊️ Gemini AI Pro የሚያቀርበው: ✅ Gemini Advanced (3.5Pro) ✅ Veo 3 – AI Video Generation ✅ Imagen 4 – AI Image Generation ✅ Anti gravity and jules agent ✅ Deep Research ✅ NotebookLM Plus ✅ 5TB Google Cloud Storage ✅ Gmail & Google Workspace Premium Features ✅ Early Access to New AI Features ...እና ሌሎች ብዙ Premium AI Features 🚀 ♊️ Gemini AI Pro – 18 ወር ▫️ Original Price: $360+ (ከ58,000 ብር በላይ) ✅ DealET ጋር በ1,500 ብር ብቻ! 🥶 📩 ለመግዛት: @dealET_support

Isn't it cheaper to plan with a larger, smarter fronteir model and execute with a cheaper one? (Yaw terra and luna fit in the cheaper bucket of course)

It’s great, I have the plus subscription. Works amazing

I was thinking to try it with plus subscription. I got Claude max and I wanna see what the gpt is capable of

📦 New Stock Added! 🎉 Product: Gemini Pro 18 months 5TB | 6 Hours HW 🛒 Quantity: 300 items ⚡️ Available now — Order while stocks last! 💵Price 500 Birr