talwinderr

Token Billionaire

[talwinderr @tlwindrr]

@heyraj__ usefullll

[talwinderr @tlwindrr]

@heyraj__ what you are buildling

[talwinderr @tlwindrr]

In LLMs the model isn't your bottleneck, your context is.

Most people dump a whole codebase or doc into the prompt and hope.
That's not context management that noise.

Real setups do 3 things:
> scope the window
> rank what's relevant
> refresh only what changed.

do this;

[talwinderr @tlwindrr]

@heyraj__ Jason not JSON

[talwinderr @tlwindrr]

@heyraj__ use sarvam

[talwinderr @tlwindrr]

@Muskanjain0401 @supermemory how i can access this

[talwinderr @tlwindrr]

@heyraj__ SECRET_KEY

[talwinderr @tlwindrr]

@heyraj__ @HelloVyom use some third party api 8$ per 1000 request i used long back

[talwinderr @tlwindrr]

@heyraj__ @XiaomiMiMo as of now you can use mimo stt & tts for free

[talwinderr @tlwindrr]

@ai_alok @Cloudflare vibe coded or you wrote code ?

[talwinderr @tlwindrr]

@rasmalai scaling, caching, ?

[talwinderr @tlwindrr]

@_heyrico linkedin is shit

[talwinderr @tlwindrr]

@craigzLiszt 😅😅

[talwinderr @tlwindrr]

@heyraj__ best of luck

[talwinderr @tlwindrr]

@heyraj__ 2026 🚀🚀

[talwinderr @tlwindrr]

@AdrianRoquell_ 1

[talwinderr @tlwindrr]

@tomcrawshaw01 Skill

[talwinderr @tlwindrr]

@ramxcodes https://t.co/tqkTmzNGW4

[talwinderr @tlwindrr]

@heyraj__ @Adityapandeydev from where

[talwinderr @tlwindrr]

@heyraj__ Mine also