📣 @GoogleAI’s Gemini 3.5 Flash is now generally available and rolling out in GitHub Copilot. Early testing shows ➡️ It has strong tool use, fast response times, and high cache efficiency ➡️ It is it well-suited for fast, iterative agentic coding workflows Try it out in @code. ⬇️ github.blog/changelog/2026…

@github @GoogleAI @code Fast tool use is great. Context continuity across sessions is better.

Gemini 3.5 Flash's tool use + caching efficiency in Copilot is solid. The "fast, iterative agentic coding workflows" framing is right—the question is whether it maintains context quality across long sessions. Tool use calibration matters enormously for code generation. Worth testing the edge cases on refactoring complexity.

@github @GoogleAI @code cache efficiency sounds boring but thats the part that decides if agents become daily workflow or just demo toy. fast iteration only matters if review and rollback stay cheap

The Wienie 500 is back. Tune in at 2pm ET on FOX to see who will taste victory, cook the competition, and earn the title of top dog.

@github @GoogleAI @code Its a lot more expensive too

@github @GoogleAI @code Fast tool use is what changes the feel of agentic coding. The model has to reason, but the workflow only works if it can iterate quickly through repo context, tool calls, tests, and human checkpoints without turning every step into latency tax.

@github @GoogleAI @code whats the real cache hit rate inside agentic loops? show me that number lowkey cache efficiency on its own, not sure it tells me real cost

@github @GoogleAI @code Fast tool use + cache efficiency is the real unlock for agentic loops. Latency compounds across iterations, this actually makes autonomous workflows viable.

@github @GoogleAI @code are u kiding me 14x

@github @GoogleAI @code x.com/i/status/20568…

@github @GoogleAI @code Iterative agentic coding isn't a model-speed problem - it's a verification problem. The faster you generate, the faster you'll commit nonsense unless there's a self-check step between tool calls. Speed without judgment just compounds bad decisions.

@github @GoogleAI @code @github just bring back Opus for students bro. Please, if you can't do it for everyone, just let it be for anyone who signed up for it before 2026 or smth, it would be cool

@github @GoogleAI @code Fast, cheap, and good at tools is exactly what coding agents need.

@github @GoogleAI @code @github help me please there's something wrong with my github account i need help. My commits are not showing and also i can't push code to github. But i can't see anything wrong with GitHub at all i can't push codes from terminal or IDE but i can with github desktop.Please help

Advertisers are using (and loving) the new X Ads Manager - no secret why - the results speak for themselves

High cache efficiency is underrated in a coding context — when you're iterating on the same codebase repeatedly, hitting cache turns repetitive context loading from a cost center into nearly free. Gemini 3.5 Flash's profile (speed + tool use + cache) is exactly right for the tight loop Copilot users run most.

@github @GoogleAI @code Why does Gemini 3.5 Flash have a higher rate even though it's cheaper than GPT-5.4?

@github @GoogleAI @code At 14x

@github @GoogleAI @code 14x!!!!

@github @GoogleAI @code high cache efficiency is the sleeper metric. Agentic coding gets expensive less from one answer and more from repeated context replay across tool calls.

@github @GoogleAI @code Is it a f joke? What the heck

@github @GoogleAI @code Fast response times don't mean a thing if the logic fails during complex agentic loops. Speed's a vanity metric when you're just hallucinating faster.

@github @GoogleAI @code A what multiplier????

3 toxic dog foods linked to paw licking. #2 is in your pantry! 🚨

@github @GoogleAI @code bom saber que tá no copilot, mas quem não usa copilot pode acessar temporariamente pelo code.verboo.ai também. não precisa esperar o rollout inteiro

@github @GoogleAI @code Good luck

@github @GoogleAI @code Bookmark this. 2026 is officially the year AI agents took over software engineering.

@github @GoogleAI @code Fix your pricing page I can't upgrade to pro

Gemini 3.5 Flash coming to Copilot is massive for day-to-day productivity. The speed and high cache efficiency make it absolutely perfect for fast, iterative code refactoring and multi-file tasks. Agentic workflows really need that low latency to feel seamless. Can't wait to test this out in VS Code today!

@github @GoogleAI @code 14X WITH LAUNCH DISCOUNT WHAT THE RUCKHAHSUAHXUAHHSHAUDBAUSİ

@github @GoogleAI @code is this available for student plan…?