Gemini 3.6 Flash quota on Antigravity is gone faster than 3.5...
Plus, 3.5 Flash Lite can't be selected on agy.
Expectations wasn't good, but wtf! Hard to support
@antigravity@GoogleDeepMind
Gemini 3.6 Flash is now live in Antigravity CLI. Weekly quota also just reset, keep building!
- more efficient tool calls per workflow
- ~17% output tokens savings compared to 3.5 flash.
- $7.50/1M output, down from $9 (3.5)
$ agy –model gemini-3.6-flash —effort medium
Have fun 🙏
Now you can use Gemma directly in the Gemini CLI! 🚀
v0.40.0 introduces experimental support for local Gemma models, starting with intelligent model routing (with full local execution on the roadmap!).
Building a fitness app that actually understands Bulking & Cutting.
Most apps: just weight loss.
My app: Intelligent Bulking & Cutting tracking with AI that adapts to YOUR goals.
GPT 5.6 is crushing it. 🚀
#FitnessApp#AI#Bulking#Cutting
@simaopedrotech@rafaquint A maioria dos modelos open weights são disponibilizados por provedores, inclusive americanos, pelo mesmo preço ou até mais baratos. Então, pelo menos na inferência (obviamente não estamos falando de P&D aqui), o modelo é rentável comercialmente sem ser subsidiado.
Building a fitness app that actually understands Bulking & Cutting.
Most apps: just weight loss.
My app: Intelligent Bulking & Cutting tracking with AI that adapts to YOUR goals.
GPT 5.6 is crushing it. 🚀
#FitnessApp#AI#Bulking#Cutting
@cline I reached the 5-hour limit only once and realized that the weekly limit has already been compromised by 40% and the monthly limit by 20%. This part is still very bad...
@jpschroeder Deepseek v4 was design from scratch to use China/Huawei hardware. Every aspect was projectes considereding cost and performance optimization.