Kirill
@kirillk_web3
AI & Web3 creator. I test AI agents and automation, report the daily AI news that matters, and keep it fun. 200M+ views.
1.1K Following 43K Followers
Anthropic CEO after someone put 29 free AI providers behind one API and dropped it on GitHub for free
🚨this is f*cking free
someone put 29 free AI providers behind one API. 37K stars on GitHub.
same idea as Free Claude Code, one level bigger.
here's what it does:
> one OpenAI-compatible endpoint for Google, Groq, Cerebras, Mistral, OpenRouter, NVIDIA and more
> ~4 billion free tokens a month across all of them
> when one provider hits a rate limit, it jumps to the next one
> tracks every key so you stay under the free caps
> works with Claude Code and Codex
one install. one key. 358 free model endpoints.
no frontier models, no uptime guarantee, and each provider's rules still apply. for side projects and testing, it's hard to beat.
repo below ↓
Show more
🚨this is f*cking free
someone put 29 free AI providers behind one API. 37K stars on GitHub.
same idea as Free Claude Code, one level bigger.
here's what it does:
> one OpenAI-compatible endpoint for Google, Groq, Cerebras, Mistral, OpenRouter, NVIDIA and more
> ~4 billion free tokens a month across all of them
> when one provider hits a rate limit, it jumps to the next one
> tracks every key so you stay under the free caps
> works with Claude Code and Codex
one install. one key. 358 free model endpoints.
no frontier models, no uptime guarantee, and each provider's rules still apply. for side projects and testing, it's hard to beat.
repo below ↓
Show more
Sam Altman’s life after this tweet:
today we launched ChatGPT. try talking with it here:
> be Anthropic co-founder
> leave OpenAI with Dario in 2021 to start Anthropic
> help build one of the most valuable AI companies on earth
> spend years warning that AI will move faster than anyone is ready for
> get invited to the IMF and World Bank meetings
> walk on stage in front of central bankers
> "One model predicts a vast economic transformation with US GDP potentially reaching $44 trillion by 2030, which sounds very outlandish, I know."
> that's almost +50% in 4 years
> then ask for serious rules for AI and the companies building it
> your company's IPO is reportedly weeks away
forecast the boom. ask for the brakes. ring the bell in November.
watch the clip below ↓
Show more
Anthropic co-founder Jack Clark:
"One model predicts a vast economic transformation with US GDP potentially reaching $44 trillion by 2030, which sounds very outlandish, I know."
US GDP today is just over $30 trillion. that's almost +50% by 2030.
think about what that actually means.
> the forecast comes from one of Anthropic's internal models
> it assumes AI keeps getting better and spreads across the economy
> his next line: "a billion virtual scientists"
> his ask: "unusual and serious rules" for AI and the companies building it
> Anthropic's IPO is expected in November, at a valuation of up to $2 trillion
the company forecasting the boom is the same one asking for restraint. weeks before its expected IPO.
which models are driving this, and what do they cost? my October 2026 guide is in my profile ↓
Show more
The most important line in Anthropic's announcement isn't the 5%.
it's this: "with an additional 2% added at the board's discretion for every 50% increase in the company's value above a set baseline."
read that again.
the bigger Anthropic gets, the bigger the share it hands to the public.
> 5% of the company goes into the fund to start
> it comes from equity that would have gone to executive pay
> co-founders pledged 80% of their own shares
> about half of employees donate shares too
> roughly 20% of Anthropic now promised to the public good
most companies tie executive pay to the stock going up.
Anthropic just tied public benefit to it.
and they announced it weeks before an IPO reportedly valued at up to $2 trillion.
Show more
We’re committing a portion of Anthropic’s shares to the Anthropic Public Benefit Fund, a new wholly owned entity dedicated entirely to the public benefit.
It will start off with 5% of our shares. For every 50% increase in the company’s value over a set baseline, we can contribute up to an extra 2%, subject to our board’s discretion.
In total, this means that roughly 20% of Anthropic’s shares are committed to the public good. Our co-founders have already pledged to donate 80% of their equity, and around half our employees are donating a portion of their equity as part of a donation matching program we’ve run since our founding.
In the post below, we share more details about our new fund.
Show more
Anthropic co-founder Jack Clark:
"One model predicts a vast economic transformation with US GDP potentially reaching $44 trillion by 2030, which sounds very outlandish, I know."
US GDP today is just over $30 trillion. that's almost +50% by 2030.
think about what that actually means.
> the forecast comes from one of Anthropic's internal models
> it assumes AI keeps getting better and spreads across the economy
> his next line: "a billion virtual scientists"
> his ask: "unusual and serious rules" for AI and the companies building it
> Anthropic's IPO is expected in November, at a valuation of up to $2 trillion
the company forecasting the boom is the same one asking for restraint. weeks before its expected IPO.
which models are driving this, and what do they cost? my October 2026 guide is in my profile ↓
Show more
Sam Altman’s life after this tweet:
today we launched ChatGPT. try talking with it here:
The most important line in Anthropic's announcement isn't the 5%.
it's this: "with an additional 2% added at the board's discretion for every 50% increase in the company's value above a set baseline."
read that again.
the bigger Anthropic gets, the bigger the share it hands to the public.
> 5% of the company goes into the fund to start
> it comes from equity that would have gone to executive pay
> co-founders pledged 80% of their own shares
> about half of employees donate shares too
> roughly 20% of Anthropic now promised to the public good
most companies tie executive pay to the stock going up.
Anthropic just tied public benefit to it.
and they announced it weeks before an IPO reportedly valued at up to $2 trillion.
Show more
We’re committing a portion of Anthropic’s shares to the Anthropic Public Benefit Fund, a new wholly owned entity dedicated entirely to the public benefit.
It will start off with 5% of our shares. For every 50% increase in the company’s value over a set baseline, we can contribute up to an extra 2%, subject to our board’s discretion.
In total, this means that roughly 20% of Anthropic’s shares are committed to the public good. Our co-founders have already pledged to donate 80% of their equity, and around half our employees are donating a portion of their equity as part of a donation matching program we’ve run since our founding.
In the post below, we share more details about our new fund.
Show more
> be Anthropic co-founder
> leave OpenAI with Dario in 2021 to start Anthropic
> help build one of the most valuable AI companies on earth
> spend years warning that AI will move faster than anyone is ready for
> get invited to the IMF and World Bank meetings
> walk on stage in front of central bankers
> "One model predicts a vast economic transformation with US GDP potentially reaching $44 trillion by 2030, which sounds very outlandish, I know."
> that's almost +50% in 4 years
> then ask for serious rules for AI and the companies building it
> your company's IPO is reportedly weeks away
forecast the boom. ask for the brakes. ring the bell in November.
watch the clip below ↓
Show more
Anthropic co-founder Jack Clark:
"One model predicts a vast economic transformation with US GDP potentially reaching $44 trillion by 2030, which sounds very outlandish, I know."
US GDP today is just over $30 trillion. that's almost +50% by 2030.
think about what that actually means.
> the forecast comes from one of Anthropic's internal models
> it assumes AI keeps getting better and spreads across the economy
> his next line: "a billion virtual scientists"
> his ask: "unusual and serious rules" for AI and the companies building it
> Anthropic's IPO is expected in November, at a valuation of up to $2 trillion
the company forecasting the boom is the same one asking for restraint. weeks before its expected IPO.
which models are driving this, and what do they cost? my October 2026 guide is in my profile ↓
Show more
Anthropic co-founder Jack Clark:
"One model predicts a vast economic transformation with US GDP potentially reaching $44 trillion by 2030, which sounds very outlandish, I know."
US GDP today is just over $30 trillion. that's almost +50% by 2030.
think about what that actually means.
> the forecast comes from one of Anthropic's internal models
> it assumes AI keeps getting better and spreads across the economy
> his next line: "a billion virtual scientists"
> his ask: "unusual and serious rules" for AI and the companies building it
> Anthropic's IPO is expected in November, at a valuation of up to $2 trillion
the company forecasting the boom is the same one asking for restraint. weeks before its expected IPO.
which models are driving this, and what do they cost? my October 2026 guide is in my profile ↓
Show more
"Claude writes about 90% of Claude Code's code."
this is dangerous
Claude writes about 90% of Claude Code's code.
From June to September, merged PRs more than doubled while the share reverted fell by more than 40%.
We wrote about how we changed code review to get there, and how to adopt it:
Show more
Nvidia CEO saying we don’t need to slow down AI
NVIDIA CEO, Jensen Huang:
"We should accelerate the development of monitoring technologies and AIs that monitor AIs."
that was his answer to why Altman, Amodei and Musk want to slow down, and he doesn't.
think about what that actually means.
> the labs building AI are asking to slow down
> the man selling them the chips wants to speed up
> his fix for dangerous AI is more AI watching it
> and the chips for both come from Nvidia
the answer to AI risk, according to the company that sells the compute, is to buy more compute.
Show more
NVIDIA CEO, Jensen Huang:
"We should accelerate the development of monitoring technologies and AIs that monitor AIs."
that was his answer to why Altman, Amodei and Musk want to slow down, and he doesn't.
think about what that actually means.
> the labs building AI are asking to slow down
> the man selling them the chips wants to speed up
> his fix for dangerous AI is more AI watching it
> and the chips for both come from Nvidia
the answer to AI risk, according to the company that sells the compute, is to buy more compute.
Show more
this is f*cking gold
the prompt that turns Claude into a motion director
if I had this a year ago, I'd have stopped paying for motion design and started shipping videos in code
in the right hands, this changes everything:
> Claude plans, animates, scores and checks a full video in code
> it locks every fact to a source before animating a single frame
> it cuts the picture to the beat, not the other way around
> a separate critic agent reviews every render until it's good enough to ship
one prompt. Remotion or HTML + GSAP. a finished film at the end.
save it and paste it into Claude tonight.
Show more
Boris Cherny built Claude Code. In 11 minutes he explains why the new Opus models need less prompting than anything before them.
and why your old detailed prompts are now working against you.
here's the part nobody is talking about:
> long instructions don't make the model smarter anymore.
> old prompt templates just add noise.
> his own team cut most of the Claude Code system prompt.
> he suggests deleting your CLAUDE.md every few months to see what the model can do without it.
most people are still prompting like it's 2025.
the model got smarter. the prompts didn't.
watch it below ↓
Show more
"Claude writes about 90% of Claude Code's code."
this is dangerous
Claude writes about 90% of Claude Code's code.
From June to September, merged PRs more than doubled while the share reverted fell by more than 40%.
We wrote about how we changed code review to get there, and how to adopt it:
Show more
Nvidia CEO saying we don’t need to slow down AI
NVIDIA CEO, Jensen Huang:
"We should accelerate the development of monitoring technologies and AIs that monitor AIs."
that was his answer to why Altman, Amodei and Musk want to slow down, and he doesn't.
think about what that actually means.
> the labs building AI are asking to slow down
> the man selling them the chips wants to speed up
> his fix for dangerous AI is more AI watching it
> and the chips for both come from Nvidia
the answer to AI risk, according to the company that sells the compute, is to buy more compute.
Show more
NVIDIA CEO, Jensen Huang:
"We should accelerate the development of monitoring technologies and AIs that monitor AIs."
that was his answer to why Altman, Amodei and Musk want to slow down, and he doesn't.
think about what that actually means.
> the labs building AI are asking to slow down
> the man selling them the chips wants to speed up
> his fix for dangerous AI is more AI watching it
> and the chips for both come from Nvidia
the answer to AI risk, according to the company that sells the compute, is to buy more compute.
Show more
🚨Claude just cracked the NSA's 26-year-old secret behind internet encryption.
it took under half an hour.
in 2000, the NSA gave the world the "random" numbers behind the NIST curves. P-256 is one of them, and it protects HTTPS, passkeys and payments today.
nobody knew where those numbers came from. the NSA engineer behind them died in 2023. there was a $36,864 bounty for anyone who could explain them.
here's what Claude did:
> guessed they came from a 1990s C random number generator
> wrote a scanner and tested 34 combinations
> spotted a pattern with a 1 in 5,810 chance of being random
> traced it to lrand48, seeded with Unix timestamps from 1997-98
the "mystery numbers" protecting the internet came from a basic random number generator and a clock.
no backdoor found. the theory that the NSA weakened the curves just got a lot weaker.
the bounty goes to the nonprofit behind Let's Encrypt.
Show more
ANTHROPIC JUST BUILT A RESEARCH AGENT LIVE. PART 2.
35 minutes. same team, now they build one from scratch.
Bookmark this for the weekend.
Skills.
Subagents.
Shared memory.
A grader that makes Claude redo the work until it passes.
Task → Rubric → Grader → Retry → Results → Money
Show more
ANTHROPIC JUST SHOWED HOW THEY BUILD AGENTS.
17 minutes. straight from the team that built it.
Bookmark this for the weekend.
One loop.
Brain split from the hands.
Sandboxes on demand.
90% faster first token.
Agents → Harness → Sandboxes → Sessions → Systems → Money
Show more
ANTHROPIC JUST SHOWED HOW THEY BUILD AGENTS.
17 minutes. straight from the team that built it.
Bookmark this for the weekend.
One loop.
Brain split from the hands.
Sandboxes on demand.
90% faster first token.
Agents → Harness → Sandboxes → Sessions → Systems → Money
Show more
