← Back to archive
Cursor doubles to $2B (in 90 days)

Photo by Markus Spiske on Unsplash

Wednesday, March 4, 2026

Cursor doubles to $2B (in 90 days)

Cursor just hit a $2B annual revenue run rate—doubling in three months (wild)—while Claude went dark amid Pentagon drama and Google Gemini finally got memory for premium users. Meanwhile MyFitnessPal scooped up teenage-founded Cal AI and its $30M in revenue. Would you pay $2B/year for AI coding tools?

Top Stories

1
AI Coding Startup Cursor Hits $2 Billion Annual Sales Rate

Bloomberg

AI coding startup Cursor has hit a $2 billion annual revenue run rate, doubling in three months with 60% coming from enterprise customers. The company's rapid growth makes it one of the most valuable AI startups at $29.3 billion, demonstrating explosive demand for AI-powered developer tools.

cursorai-codingfundingenterprise-ai-adoption
2
Anthropic Claude Experienced Widespread Outage

TechCrunch

Claude experienced major outages affecting login and core services during a traffic surge driven by heightened attention from its Pentagon negotiations controversy, which propelled the app to the top of the App Store charts above ChatGPT.

anthropicclaudeoutagechatgpt
3
Google's Gemini chatbot now has memory

TechCrunch

Google's Gemini chatbot now offers a memory feature for premium subscribers that stores user preferences and context, similar to ChatGPT's capability, though security concerns around such features persist as they can be exploited by malicious actors.

geminigooglechatbotllm
4
MyFitnessPal acquired Cal AI

MyFitnessPal acquired Cal AI, the viral AI calorie-counting app built by teenage founders that achieved $30 million in annual revenue and 15 million downloads in under two years. The deal allows Cal AI to operate independently while leveraging MyFitnessPal's extensive nutrition database, targeting users who prioritize speed over detailed tracking.

acquisitionconsumer-aicomputer-visionhealthtech
5
Just-in-Time Agentic Memory Framework

arXiv

A new 'just-in-time' memory framework called GAM dynamically generates optimized contexts for AI agents at runtime rather than relying on static pre-loaded memory, achieving significant performance improvements by leveraging LLM capabilities and reinforcement learning.

agentsllmmemoryreinforcement-learning

Keep Reading

Industry Voices

Enjoyed this issue?

Get daily AI intel delivered to your inbox. No fluff, just the stories that matter.