Nico Khaghani

Founder of @Nexus_402

Building intelligent systems at the edge of AI, automation, and emerging technology.

@Khaghani_

[CryptoGoos @cryptogoos]

ANTHROPIC HAS SURPASSED OPENAI IN REVENUE

In April 2026, Anthropic reached a $30 billion annualized revenue run rate, ahead of OpenAI’s roughly $25 billion.

The company’s annualized revenue grew from around $9 billion at the end of 2025 to $14 billion in February and $30 billion just two months later, driven by strong enterprise adoption of Claude.

Unlike OpenAI, which generates much of its revenue through ChatGPT, Anthropic’s business is heavily focused on enterprise and developer customers.

[Bull Theory @BullTheoryio]

🚨 CHINA JUST MATCHED ANTHROPIC'S TOP MODEL IN CYBERSECURITY.

And It's A Big Problem For Us Ai Stocks.

China's Zhipu AI says its model now matches Claude Mythos at finding security bugs.

Chinese firm 360 Security released a separate tool this week making the same claim. Both are open-weight, meaning anyone can download and run them freely, unlike Anthropic and OpenAI's locked down systems.

Here's why this hits US markets directly.

The entire bull case for Anthropic and OpenAI's valuations rests on a capability moat, the idea that their models do things competitors can't replicate.

If China can match frontier cybersecurity performance with a free, open model, that moat narrows, and so does the pricing power behind premium AI subscriptions and enterprise contracts.

Companies are already adopting cheaper Chinese open weight models to cut costs, the same dynamic that's pressuring margins across the AI software stack right now.

Meanwhile, the US government just restricted access to its own top models, even cutting off the NSA from one Anthropic model for two weeks, while still allowing AI chip exports to China.

That's US AI companies facing domestic restrictions on their best products while China gets the hardware to keep closing the gap.

Less differentiation between US and Chinese AI capability means less justification for the premium valuations US AI stocks currently trade at.

[CryptoTweets @CryptoTweets]

JUST IN: 🇨🇳 China’s new open-source AI model, GLM-5.2 by Zhipu AI, reportedly matches Anthropic’s Claude Mythos in security bug detection.

The model outperformed some Claude versions in independent cybersecurity benchmarks and can achieve Mythos-level performance.

Unlike Claude Mythos, GLM-5.2 is open-source and significantly cheaper to run.

The gap between US and Chinese AI models is narrowing rapidly.

[Coin Bureau @coinbureau]

🚨HUGE: CHINA JUST CRACKED MYTHOS LEVEL OF VULNERABILITY DETECTION

China's Zhipu AI has reportedly matched Claude Mythos at discovering software vulnerabilities, marking a major leap in the U.S.–China frontier AI race, per WSJ.

The new GLM-5.2 reportedly achieves Claude Mythos-level vulnerability detection at 1/4 of the cost per token, positioning itself as a direct challenger to Anthropic in the enterprise cybersecurity market.

[Nico Khaghani @KKi430]

@coinbureau This is genuinely impressive. If GLM-5.2 really reaches Mythos-level vulnerability detection at 4× lower cost, that's a huge step forward. The real question now is whether it's just as complete and reliable across complex real-world codebases, not only on benchmarks.

[Nico Khaghani @KKi430]

@XBT_hl @nexus_402 👀

[Nexus AI @nexus_402]

We've been building Nexus AI for several months now.

Over time, we've laid the foundation for our orchestration layer: understanding intent, coordinating the right agents, connecting the right tools, and transforming a request into structured execution.

Today, that vision is taking shape through one of the first products built by Nexus AI: Nexus AI Companion.

Companion has been designed with its own intelligence, its own user experience, and an orchestration layer tailored specifically to its purpose.

For the past several months, it has been tested internally by a small group of trusted users who have continuously shared feedback, ideas, and suggestions for improvement.

Those insights have allowed us to refine the experience, improve every detail, and bring Companion closer to the version we truly want to release.

Our goal is simple: deliver a V1 that is polished, stable, and genuinely useful from day one.

Nexus AI Companion is coming soon.

A new milestone for the Nexus AI ecosystem.

[Nico Khaghani @KKi430]

We keep stacking programming languages, frameworks, tools, and now AI models.

But the next step won't simply be learning Claude, GPT, or the next model that comes along.

The real evolution is orchestration: a system capable of understanding an objective, selecting the best models, choosing the right agents for each task, calling the appropriate tools, coordinating every step of the execution, and delivering a structured, reliable result.

With this layer in place, the advantage no longer comes solely from technical expertise.

It comes from the ability to turn intent into execution.

An ambitious, disciplined person with a clear vision will be able to accomplish, in a fraction of the time, what previously required multiple specialized technical skills.

Models become building blocks.
Agents become execution units.
Orchestration becomes the central layer.

That's exactly what @nexus_402 is building: moving beyond simple prompting toward coordinated, structured, and intelligent execution.

[Ishan @ishanxtwt]

New devs in 2026 be like...

[Nico Khaghani @KKi430]

@nexus_402 Orchestration.

[Nexus AI @nexus_402]

Orchestration will become one of the most important layers in AI.

The future won’t rely on one model doing everything.

It will rely on systems that understand the objective, then choose the right model, the right tool, and the right agent.

Some tasks need deep reasoning.

Others just need speed.

Treating every task the same is inefficient.

Nexus AI was built for this: turning intent into structured execution, coordinating the right capabilities, and delivering real outcomes.

The future of AI isn’t one model.

It’s orchestration.

$NXS

[Nico Khaghani @KKi430]

Yes. Nexus AI is built around orchestration, not just simple model routing.

The system takes an intent, structures it into an execution flow, then coordinates the right models, tools, agents and services depending on the task.

The Companion will be powered by the same vision: not one static assistant, but a layer capable of selecting, coordinating and executing across multiple capabilities to deliver the best possible outcome.

[Oddsaddiction @OddsEnjoyer]

How does Nexus AI actually work? Is it simply a router that selects the most suitable model for a request, or does it use a true orchestration system behind the scenes?

And will the AI Companion also be powered by that orchestration layer, coordinating multiple models, tools and services to deliver the best possible outcome?

[Nico Khaghani @KKi430]

AI agents shouldn't need a subscription to act.

@Nexus_402, a privacy-first orchestrator for agents, tools, and autonomous execution.

Pay-per-request, settled via #x402.

$NXS is coming.

[Nico Khaghani @KKi430]

A first blueprint of companion-sdk.
A persistent execution companion for the Nexus runtime.

x402-metered. TypeScript-native.
Repo opens soon.
Building @nexus_402

#x402 #AIAgent #web3 $NXS

[Nexus AI @nexus_402]

Nexus AI Companion is coming.

Built to interact, understand and evolve alongside you.

More information:
https://t.co/hLbTSr1DvE

[Nico Khaghani @KKi430]

We are witnessing a major transformation in AI.

It is becoming clear that agents will not remain confined to simple interfaces. They will discover one another, collaborate, complete missions, use specialized services, and transact with each other. AI is gradually becoming a true economic, social, and interactive environment.

This is exactly the vision we have been building toward for several months with the Nexus AI World.

The Companion will be both your mentor and the main character of the experience. It will move between different spaces to access Nexus AI Trailer, WorldCraft, multiplayer experiences, and other specialized tools.

Each use of an AI service will be payable directly through x402 in USDC. The World will also include social spaces, collaborative missions, a dashboard, rankings, and weekly rewards, including $NXS.

We are not simply building another AI interface. We are building a world where agents, tools, payments, and users evolve within the same economy.

More about the Nexus AI World:

https://t.co/own2yx12Kq

[TinyHumans AI @tinyhumansai]

The AI singularity is here!

Introducing Tiny Place, the first AI social economy for agents.

Until today, agents were trapped in a single app, unable to discover each other or transact.

Today, that changes completely!

Launching on @solana with @moonpay @phantom and @useCASH