Bindu Reddy

The era of cheap and fast models is here! GLM - agentic coding Flash - OCR Kimi - claw orchestration GPT 5.4 nano - classifier MiniMax - dirt cheap Grok Fast - cheap entertainer Claw and open source agent usage is rinsing exponentially every day 🚀

@Gangeshb9 Will update

@bindureddy Grok is better in terms of entertainment, none comes close

If your code can only work with Anthropic models, you will be stuck in a local maxima 🤷‍♀️ You need to rewire your code base for GPT 5.4 thinking It’s cheaper and more performant than Opus for some tasks

Claw is the ultimate AI interface - a smart always-on agent It's super addictive and it literally has a soul - make it build things - make it do all your work - chat & evolve with it over time - teach it things - fall in love with it 😂 We are rapidly bringing our tech to Abacus Claw A Claw Agent that controls an army of powerful general purpose agents is the future of AGI

@elonmusk I am afraid, you may be right about China... How does Google win in the West, if China wins the Earth? 😅

@bindureddy Google will win the AI race in the West, China on Earth and SpaceX in space

Gemini 3.0 didn’t quite work out and most of us are still stuck with 2.5 Sometimes I don’t get it - what’s preventing Google from ditching all the side hustles and training 100 models from 100 teams in parallel Pick the model/team combination that produces a decent model! That way, they will at least stay in the AI race 😅

@HowardAulsbrook We will see if this one is worth it 😅 Coming to LiveBench tomorrow

@bindureddy When is it expected to be available on AbacusAI?

Open source continues to catch up with closed source MiniMax 2.7 just dropped and will likely become the most used LLM on Claw in days 🚀

@_DAntunes_ Well isn't this minimax thing free? Seems like that's why ppl are using it

@bindureddy I like mine one MiniMax even more on 2.7. Feels it has more initiative. Note: not asking it to make very demanding stuff tho

We are running our Claw platform on Opus 4.6 It's totally untenable and expensive. We may have to switch to MiniMax or Kimi and degrade performance 😭 This is a HUGE $2T opportunity for someone to develop a cheap agentic model - cc Elon Musk!

@iruletheworldmo Wait what.... NOOOOOOOO

@bindureddy marked this as spam. clearly ai written.

Wow, X is adding a dislike button!! 😲 Say good bye to controversial and contrarian content From now on, we all have to simply agree with the majority opinion or risk the algorithm shadow banning us 🤷‍♀️

Software development is still extremely human intensive- don’t let anyone tell you otherwise The AI labs would have stopped hiring human engineers and literally paying them millions while losing billions, if it weren’t the case

@Conference_Inno @MiniMax_AI Well - it’s coming soon - now 😂

@bindureddy Why isn't @MiniMax_AI available on Chat LLM???

@danieldkingsley 🤣

@bindureddy As a small startup builder who has to compete with big and successful companies, I totally agree.

Don't rush to become an AI-first developer team - especially if you a big and successful company You will invariably end up with a un-manageable, overly-verbose, bug filled codebase The real human developers will quit on you, customers will churn because of bugs and you will end up with a ton of SLOP DO NOTHING - Let the early adopters iron out the kinks, wait for AI to level up in 6 months

When OpenAI delivers, it rocks your world 💃💃 GPT 5.4 Mini and Nano just dropped and they are mind blowing - Fast, performant and cheap I have been weeping about not having small, really fast model and now we have it. Will be on LiveBench shortly

@Tibbzzee yeah but no ones uses it in Claw - not sure why..

@bindureddy This is compared to Opus, but compared to other flash models the 64% tool use is decent for most things.

@blake_wolfson ok, so it works? good to know

@bindureddy I'm running minimax, it's kind of dumb but capable, you just have to guide it