Dragos

@Hangsiin Is better then Gemini

DeepSeek V4 Flash has become the new Gemini Flash.

@TeksEdge @OpenRouter It is cheaper then a certain flash model :)

Great news! Qwen3.7 Max is now on @OpenRouter @ $7.5/1M and decent output at 96 tps.

@twentythrill This is gemini 3.5 pro. This is how they doit now, Google moves backwards now :)

Gemini 3.1 Pro is completely broken since the platform update. We’re paying more for tighter compute limits, only for the model to hallucinate wildly and completely fail to read basic chat context. I don't know what they did to the backend, but they massacred my boy.

Unpopular opinion? Gemini 3.5 Flash feels like a regression over 3.1 for my use cases. Hard to justify when it’s also 3x more expensive.

@ai_for_success Yes but is dumb as a rock :)

The best place to use Gemini 3.5 Flash is Antigravity CLI. If you haven’t installed it yet, it’s pretty simple to set up and get started. And now the rate limits are 3x higher :) which is amazing.

Me: "I need Claude Pro for coding" My wallet: "You need OpenCode Go" Me: *codes with 12 Chinese AI models for $5/month* My wallet: "Finally, a developer who listens" youtube.com/watch?v=XGQIBn…

@LexnLin Yessssss

Is Deepseek v4 flash/pro better than Gemini 3.5 Flash?

@screenstudio AI mistake :)

Update on the subscription cancellation emails: We've already resolved this with our payments operator, Lemon Squeezy. It turns out it happened by accident on their side due to a support agent mistake. Their team is already handling it, and your subscription should be restored very soon. Your access to the Screen Studio app should stay uninterrupted in the meantime. We're sorry for the inconvenience and confusion. If you have any questions, we're happy to help.

@Tiny_Fish Exa and firecrawl using them both :)

500 free credits to whoever identifies the 2 competitors 😆

@zeddotdev @baseten Yesssssss

Would you buy Zed Pro if we added open source models from our friends at @baseten

@0xSero Droid is very good. It is just runs better. Tried glm 5.1 in opencode and in droid and droid just use it well it gets the code and what to do. Try it and see

I think it’s time for everyone to give Droid a shot. New usage system. - 100$ plan - usage pools - secondary usage pools with Droid Core (open weight models) I love this

@mistralvibe Sounds good

📰 Some news for you tomorrow. Big news.

@kentcdodds Than you uavent seen this :)

I'm currently writing an article titled "The Last Software Engineer"

This is vibe coding and the dangers if you don’t have a background in software engineering. Not everyone can pilot a plane :) Probably GPT-6 will be better.

@jayair Who is first?

Update on DeepSeek, it’s already second in usage on OpenCode Go

@lassevjl I have moved the pannels back, after enabling it :)

Let's goo!

@kimmonismus With GPT or Opus?

75% of all new code at Google is now AI-generated and approved by engineers, up from 50% last fall. 2027 90%, and 2028…?

@getdokploy really just added an AI log analyzer so I finally have a robot to stare at my Docker errors for me while I pretend to know what 'exit code 1' means. Work smarter, not harder! 🤖🚢😂 Check out the full breakdown here: youtube.com/watch?v=idGJo9…

RIP Nano Banana 2. You served me well, but you also occasionally made me look like a melted plastic action figure. 🍌🪦 Enter: GPT Image 2. 👑 I put both models to the ultimate test to see who generates the best YouTube thumbnails. Watch the full showdown here: youtu.be/XlpmPzEjPAw

@webdevcody No. GPT-5.4 and GLM 5.1

does anyone actually use Gemini for coding?