K
Koda Intelligence
boltThe Signal

Intelligence Report • 19 April 2026

The Signal

Anthropic crossed $30 billion in revenue the same week Iran fully sealed the Strait of Hormuz.

Lead Story Consolidation The Decoder via LLM Stats

Anthropic Revenue Tops $30B Annualized

Anthropic has reportedly flipped from money-loser to revenue powerhouse in just a few months. Annualized revenue now tops $30 billion, possibly edging out OpenAI, and investors are already floating a potential trillion-dollar valuation for the company.

Continue Reading arrow_forward
smart_toy AI

Anthropic's annualized revenue has surpassed $30 billion, cementing its position alongside OpenAI as one of only two AI labs generating revenue at the scale of major enterprise software companies.

public World

Iran fired on commercial tankers and fully closed the Strait of Hormuz, prompting Trump to threaten destruction of Iranian infrastructure while diplomatic talks stall over what Tehran calls 'maximalist' U.S. demands.

trending_up Markets

Fear dominates as the Hormuz closure threatens roughly 20% of global oil transit, sending energy prices surging and dragging risk assets lower despite blockbuster AI earnings.

bolt Wild Card

China's new regulations on AI digital humans signal that Beijing is racing not just to build frontier models but to govern synthetic identities before they become indistinguishable from real citizens online.

Market Snapshot

query_stats

S&P 500

7,126.06

+1.20%

Nasdaq

24,468.48

+1.52%

Bitcoin

$$75,853.31

+0.16%

ETH

$2,331.13

-0.85%

Crude Oil

$90.38

-7.94%

Fear & Greed

27

Fear

Today's Focus

01

Anthropic's $30B Milestone

Anthropic's annualized revenue topping $30 billion marks a dramatic acceleration from roughly $2 billion just 18 months ago, driven by enterprise API demand and Claude's expanding footprint. The figure validates the company's safety-first positioning as commercially viable, not just philosophically aspirational. It also intensifies the funding arms race, as competitors must now match both capability and revenue trajectory to attract top-tier capital.

02

Hormuz Fully Closed

Iran's decision to fire on tankers and seal the Strait of Hormuz represents the most severe disruption to global energy flows in decades. With diplomatic channels frozen, the crisis is already rerouting international shipping, spiking insurance premiums, and forcing airlines to cancel Middle East overflights. The knock-on effects reach AI directly: energy price shocks raise the cost of powering data centers at precisely the moment compute demand is exploding.

03

Open-Source Giants Emerge

Zhipu AI's release of the 744-billion-parameter GLM-5.1 as open source represents China's most aggressive move yet to democratize frontier-scale models outside Western ecosystems. Combined with GPT-5.4 leading professional benchmarks on the closed side, the gap between open and proprietary models is narrowing faster than most forecasters expected. Security researchers warn that AI agent proliferation, fueled by both open and closed models, is creating attack surfaces that current defenses cannot match.

Listen & Watch

Daily Broadcasts
Daily Deep Dive
Podcast ~22 min

Daily Deep Dive

Video Briefing
play_arrow
Video

Video Briefing

Daily Infographic
Infographic

Intelligence Map

The Wire

AI Intelligence
Benchmark

GPT-5.4 Leads Benchmarks Across Professional Tasks

OpenAI's GPT-5.4, released March 5, continues to dominate professional benchmarks with scores of 83% on GDPval, 73.3% on ARC-AGI-2, and 91% on BigLaw. It ties with Google's Gemini 3.1 Pro at 57 on the AAI Composite index, though Gemini leads on...

Kers AI

Open Source

Zhipu AI Open-Sources 744B GLM-5.1 Model

Chinese lab Zhipu AI released GLM-5.1, a 744 billion parameter mixture-of-experts model for text and reasoning, under an MIT open-source license. Priced at roughly $1 per million input tokens and $3.20 per million output tokens, it includes the...

WhatLLM

Model Release

Microsoft Ships Three Foundational MAI Models

Microsoft launched three new foundational models on April 2: MAI-Voice-1 for voice generation, MAI-Image-2 for image and video generation, and a text model. All are available through Microsoft Foundry and the MAI Playground, developed by the...

TechCrunch

Trend

April Model Tracker Surpasses 274 Tracked Models

The AI model ecosystem continues its explosive expansion, with trackers now monitoring over approximately 274 models as of mid-April 2026. The pace of releases has forced engineering teams into constant adaptation cycles, with developers...

SearchCans

Trend

Prediction Markets Favor Grok 5 and DeepSeek V4

Manifold prediction markets give GPT-5.5 an 83% chance and Grok 5 (or any Grok model) a 79% chance of shipping before April 30. DeepSeek V4 sits at 67% and Minimax M3 at 60%, while Claude Mythos has already resolved as shipped. A GPT-6 release...

Manifold Markets

The Globe

World Affairs
Coding

Claude Code: Run Autonomous Refactors Across Large Codebases From Your Terminal

Claude Code operates as a terminal-native agent with a 200K token context window, letting you point it at an entire repository and issue plain-language refactoring instructions. Use it for cross-file dependency updates, test generation, or migration tasks where IDE-based copilots lose context. Pair it with git branch isolation so you can review diffs before merging autonomous changes.

Productivity

Criteo GO: Enterprise-Grade Ad Optimization Now Accessible to Solo Founders

Launched in April 2026, Criteo GO uses generative AI to automate ad creative generation, audience targeting, and bid optimization across retail media networks. Startups with limited marketing budgets can connect their product catalog and let the system generate and A/B test display ads without a dedicated media buyer. Start with a single product line to benchmark ROAS before scaling spend.

Build

AgentGPT: Spin Up Goal-Oriented Autonomous Agents in Your Browser

AgentGPT lets you define a high-level objective, then deploys a chain of sub-agents that research, plan, and execute tasks autonomously. Use it for competitive analysis, content outlines, or market research where you need breadth across dozens of sources quickly. Set explicit scope boundaries in your initial prompt to prevent the agent from drifting into irrelevant subtasks.

Coding

Cursor Composer 2: Orchestrate Multi-File Edits With an AI-First IDE

Cursor Composer 2 ranks as the top-rated AI coding environment for complex, multi-file projects as of April 2026. Its agentic workflow mode lets you describe a feature in natural language and watch it scaffold changes across components, routes, and tests simultaneously. Use the inline diff review to accept or reject each change granularly before committing.

Mindset

Audit Your Agent Stack: Most Teams Run Overlapping AI Tools Without Knowing It

With over 50,000 AI tools on the market and new agent builders launching weekly, redundancy is the default. Before adding another coding agent or workflow automator, map every tool your team touches to a simple capability matrix: code generation, code review, task automation, content creation. Kill any tool whose primary function is already covered by another with 80% overlap.

Creativity

Rosebud AI: Generate Game Assets and Interactive Scenes From Text Prompts

Rosebud AI, trending at +1,200% growth on Exploding Topics this month, lets creators describe characters, environments, or game mechanics in plain language and receive playable prototypes. Use it to rapidly test game concepts or generate consistent 2D and 3D assets for indie projects without a dedicated art team. Export assets directly into Unity or Unreal Engine pipelines.

The Arena

Competitive Intel

OpenAI

Expands enterprise infrastructure with Containers product and Safety Fellowship program

Meta AI

Capital expenditure plan for 2026 reaches $115-135 billion, nearly doubling year-over-year

Anthropic

Claude Mythos 5 positions Anthropic at the frontier with 10 trillion parameters

Never miss a Signal.

Join intelligence professionals who receive our daily deep-dive into AI, geopolitics, and markets.

Daily intelligence. No spam. Unsubscribe anytime.