Sabitlenmiş Tweet
I cut my Claude token bill by ~30% while still using AI daily.
Built a tiny local tool called AI Badger for exactly this. Full story in the Article below 👇
English
Ted Kupolov
418 posts
Ted Kupolov
@kupolov
Building AI Badger: Local codebase context for any LLM. Cut token burn & rate limits. Ex-Google/Apple → https://t.co/IwFwn56wca
San Diego, CA Katılım Ağustos 2009
62 Takip Edilen31 Takipçiler
@xixihk11 @signulll AI Badger - paste diff, get review, no token burn. Works with any AI chat. Open source.
pvrlabs.xyz/aibadger
English
@SinaHartung Google has two co-founders. One was born in America, the other arrived at age six. Like in other examples, country of birth had very little to do with success - it's about the opportunities the US gives legal immigrants.
English
Google: Sergey Brin (Russia 🇷🇺)
NVIDIA: Jensen Huang (Taiwan 🇹🇼)
Tesla / SpaceX: Elon Musk (South Africa 🇿🇦)
Intel: Andy Grove (Hungary 🇭🇺)
Yahoo: Jerry Yang (Taiwan 🇹🇼)
Zoom: Eric Yuan (China 🇨🇳)
Stripe: Patrick & John Collison (Ireland 🇮🇪)
Instacart: Apoorva Mehta (India 🇮🇳)
Datadog: Olivier Pomel (France 🇫🇷)
DoorDash: Tony Xu (China 🇨🇳)
Sina@SinaHartung
the elites don’t want you to know this but SF runs on immigrants
English
@sflorimm Learing Claude, like in "Learing Center"? But on a serious note, learning Claude seems more like taking HTML courses during the dot-com days.
English
The thing I don't get is:
Claude Code is 100% self-written code,
But Anthropic has 400+ engineering job openings?
English
@rezoundous I hope this motivates them to make Copilot better. When I ran into Claude throttling, I created AI Badger - to save on tokens and offload some of the agentic work to external AI chats.
pvrlabs.xyz/aibadger
English
@theo IDEs are the last line of defense to stay technical. I keep IntelliJ on standby for quick diffs and edits. To save tokens I outsource brainstorming to external chats with AI Badger.
pvrlabs.xyz/aibadger
English
You guys still use IDEs?
DROID@droidbuilds
developers, which IDE genuinely has the potential to replace VS Code?
English
@NickSpisak_ @ChatGPTapp @OpenAI Try AI Badger. It's like takeout for your codebase context - brainstorm and refine without burning local agent tokens.
How I cut my token usage by ~30%:
x.com/kupolov/status…
Ted Kupolov@kupolov
English
CODEX LIMITS NUKED TODAY
Went through all Pro usage in less than a day
Gotta wait 5 days until reset….
Something seems off @ChatGPTapp @OpenAI
English
This is the equivalent of saying, given infinite time, monkeys on typewriters would eventually write Shakespeare:
Peter Steinberger 🦞@steipete
People freaking out over my AI spend. What nobody sees: Part of what excites me so much about working on OpenClaw is that I'm trying to answer the question: How would we build software in the future if tokens don't matter? We constant run ~100 codex in the cloud, reviewing every PR, every issue. If a fix on main lands, @clawsweeper will eventually find that 6 month old issue and close it with an exact reference. We run codex on every commit to review for security issues (as it's far too easy to miss). We run codex to de-duplicate issues and find clusters and send reports for the most pressing issues. We have agents that can recreate complex setups, spin up ephemeral crabbox.sh machines, log into e.g. Telegram, make a video and post before/after fix on the PR. There's codex that watch new issues and - if it fits our documented vision well, automatically create a PR of it. (that then another codex reviews) We have codex running that scans comments for spam and blocks people. We have codex instances running that verify performance benchmarks and report regressions into Discord. We have agents that listen on our meetings and proactively start work, e.g. create PRs when we discuss new features while we discuss them. We build clawpatch.ai to split all our projects into functional units to review and find bugs and regresssions. We do the same split for security with Vercel's deepsec and Codex Security to find regressions and vulnerabilities. All that automation allows us to run this project extremely lean.
English
@Jason My 10x $BTC thesis is that altcoins will go to zero, and their value will flow into BTC.
English
@ziwenxu_ @thsottiaux Looks like a bug. When Codex starts a new session, usage drops to 80% - haven't seen that before. I offload reviews and brainstorming to AI Badger (external AI chats, no token burn)
pvrlabs.xyz/aibadger
English
Is something broken with Codex?
The usage is draining insanely fast @thsottiaux
It's only been 1 day and I'm already at 23%
I only have 2 /goals running and I have pro plan
English
@AliceBunnyland2 In CA, even the engineers' dogs make 70k after taxes. Plenty of professions are paid well above 100K post-tax: doctors, even dentists.
English
@0xRemakw @thsottiaux I've only noticed that context consumption is much faster now. But at least with AI Badger you can save 30% or more on tokens. Article below:
x.com/kupolov/status…
Ted Kupolov@kupolov
English
CODEX LIMITS REDUCED 50%
can someone verify if this is true ? @thsottiaux
English
@wesbos And if $300/month is too steep, try AI Badger with Grok Web. It saves tokens on local coding tasks while still using Grok's reasoning - perfect for code reviews and brainstorming.
pvrlabs.xyz/aibadger
English
@jturntdev You don't need high reasoning for well-scoped execution. I use AI Badger to extract relevant context, then brainstorm in any external AI chat with high reasoning enabled - without burning Codex/Claude tokens on context.
pvrlabs.xyz/aibadger
English
