It’s Monday, February 16th: Google shipped a major upgrade to Gemini 3 Deep Think that dominates every reasoning benchmark in sight, MiniMax dropped frontier-level coding performance at 1/20th the cost of Claude, and Anthropic closed a $30 billion round at a $380 billion valuation as Claude Code hits $2.5 billion in annualized revenue.
Head over to our Events Portal to get the latest on upcoming AI Collective events near you. Search by city, date, or event format, and join thousands of builders at events across 100+ chapters on every continent (except Antarctica, for now).
🌁 Based in SF? Check out SF IRL, MLOps SF, GenerativeAISF, or Cerebral Valley’s spreadsheet for more!
The AI news you need to know: weekly 🤝 by AJ Green
⚡ Google’s Deep Think Crushes Every Reasoning Benchmark in Sight
News: Google just shipped a major upgrade to its Gemini 3 Deep Think reasoning mode, posting dominant scores across math, coding, and science while simultaneously unveiling Aletheia, an Olympiad-level math research agent powered by the new upgrade.
Details:
Deep Think hit 84.6% on ARC-AGI-2, a 16-point gap over Opus 4.6 (68.8%) and 32 points above GPT-5.2 (52.9%), and set a new record of 48.4% on Humanity’s Last Exam.
It reached gold-medal performance on the 2025 Physics & Chemistry Olympiads and scored 3,455 Elo on Codeforces, nearly 1,000 points above Opus 4.6.
Google introduced Aletheia, a math agent that autonomously solves open problems, verifies proofs, and generated a research paper without human intervention, scoring 95.1% on IMO-ProofBench Advanced.
The upgrade is live now for Google AI Ultra subscribers in the Gemini app, with API access available to researchers through an early access program.
Why It Matters: After Google dominated benchmarks to close out 2025, the spotlight shifted hard toward Anthropic and OpenAI in the new year. This upgrade is a forceful correction. Deep Think’s 84.6% on ARC-AGI-2 is a GENERATIONAL gap over the next closest competitor. Aletheia signals something bigger: we’re entering the era where AI agents don’t just solve known problems, they discover new mathematics. Hats off to Google, well played.
💰 MiniMax’s Open-Source M2.5 Hits Frontier Coding Performance at 1/20th the Cost
News: Chinese AI lab MiniMax launched M2.5, an open-source model rivaling Opus 4.6 and GPT-5 on agentic coding benchmarks at a fraction of the cost, dropping the price of the frontier by as much as 95% and making it cheap enough to run AI agents continuously around the clock.
Details:
M2.5 shows particularly strong coding performance, scoring 80.2% on SWE-Bench Verified and roughly even with Opus 4.6 (80.8%) and GPT-5.2 (80.0%) across key development benchmarks.
Two API tiers are available: M2.5-Lightning ($2.40/M output tokens) and standard M2.5 ($1.20/M output tokens), both priced dramatically lower than Opus at $25/M.
MiniMax reports that M2.5 now handles 30% of the company’s daily tasks across R&D, product, sales, HR, and finance, plus 80% of all new code commits.
The first open model to exceed Claude Sonnet on independent benchmarks, though open-source weights and licensing details have not yet been published.
Why It Matters: Every few months, a Chinese lab drops a model that rewrites the cost equation for the entire industry. M2.5 delivering frontier-level coding at $1.20/M output tokens vs. Opus’s $25/M isn’t just a pricing story: it’s a structural shift in how companies deploy AI agents for long-running, autonomous tasks. Effectively, you could run your entire agentic stack for $1/hr, or $720/mo. Not running on demand, but continuous 24/7 runtime. If you’re running agents on extended workflows and paying top-tier rates, M2.5 just changed your cost model by 20x. When intelligence gets this cheap, the calculus flips from “can we afford to use AI?” to “can we afford not to run agents 24/7?”
💰 Anthropic Closes $30B Series G at $380B Valuation, With Claude Code Driving a Revenue Explosion
News: Anthropic closed a $30 billion Series G at a $380 billion post-money valuation, more than doubling its worth since September and marking the second-largest private tech financing deal of all time. The round, led by GIC and Coatue, comes as the company’s run-rate revenue hits $14 billion, with Claude Code alone generating over $2.5 billion in annualized revenue.
Details:
The round was co-led by D. E. Shaw Ventures, Dragoneer, Founders Fund, ICONIQ, and MGX, with participation from Sequoia, Lightspeed, Jane Street, the Qatar Investment Authority, and previously announced commitments from Microsoft and Nvidia totaling up to $15 billion.
Anthropic’s run-rate revenue has grown 10x annually over the past three years, with customers spending $100K+ annually growing 7x in the past year and 500+ organizations now paying at least $1 million per year.
Claude Code’s $2.5B run-rate has more than doubled since January 1, 2026, with business subscriptions quadrupling over the same period and enterprise users representing more than half of Claude Code’s total revenue.
Eight of the Fortune 10 are now Claude customers, and 80% of Anthropic’s business comes from enterprises, a key differentiator from OpenAI’s consumer-heavy model.
The capital will fund Anthropic’s $50 billion infrastructure initiative, including data centers in Texas and New York coming online later this year, built in collaboration with Fluidstack and AWS.
Why It Matters: From its first dollar in revenue less than three years ago to a $14 billion run rate, Anthropic is one of the fastest scaling companies in tech history. But the real story is Claude Code: a single product generating $2.5 billion, doubling in six weeks, with enterprise subscriptions quadrupling. That’s not a feature. That’s a platform shift. Ramp data shows 1 in 5 businesses now pay for Anthropic (up from 1 in 25 last year), and 79% of OpenAI customers also pay for Claude. Nowadays, 8 out of 10 of the Fortune 10 are being built with Anthropic... a year ago there was no clear winner of the enterprise market... oh how the times have changed.
Your pulse on the biggest events and announcements and happening in AI this week.
📅 Events We’re Watching
This is a big week for announcements, with a major conference already under way. It’s the India AI Impact Summit, the biggest AI governance gathering yet in the Global South and the headliner of this week.
Here are a few things to watch for:
February 16 – 20: India AI Impact Summit (New Delhi, India)
More than 35,000 attendees from 100+ countries are expected, including top AI lab CEOs, heads of government, and the UN Secretary-General. The agenda is focused on AI safety, governance, and deployment at scale across the developing world.
February 18 – 20: DeveloperWeek 2026 (San Jose, California)
Billed as the world’s largest developer and engineering conference and expo, bringing together 5,000+ developers and tech leaders at the San Jose Convention Center plus online.
🔦 Spotlight On: The AI Commercialization Match Heats Up
For the first time, the marquee global AI summit is being held not in London, Seoul, or Paris, but in New Delhi. The India AI Impact Summit is framed around “Impact,” indicating a less-than-subtle shift from high-level safety conversations toward concrete deployment, especially in health, education, and public services in the Global South. India is positioning itself as a bridge between advanced AI economies and emerging markets, leveraging its rapidly growing AI ecosystem.
The Paris AI Action Summit’s joint declaration last year was not signed by the U.S. or UK, exposing real fractures in how major powers view AI rules. New Delhi is now trying to broker a more inclusive consensus, with working groups on human capital, safe and trusted AI, and access to compute running alongside leader-level sessions. We’re keeping our eye on whether the summit’s Leaders’ Declaration attracts broad sign-on from major players and the wider Global South.
📝 Community Notes
🥵 GOING FAST! Get Your AIC Discounted Pass to DeveloperWeek 2026 (Feb 18-20)
DeveloperWeek 2026 runs Feb 18-20 in San Jose, bringing together thousands of developers for sessions on AI, cloud, and DevOps. Produced by DevNetwork.
A limited number of FREE OPEN Passes are still available (normally $195). Claim yours before they’re gone!
🌁 NEARCON 2026 (Feb 23-24)
The AI Collective is partnering with NEARCON 2026, a two-day gathering exploring how AI and open infrastructure are converging to power the next generation of applications, markets, and digital experiences.
NEARCON takes place Feb 23-24 in San Francisco. The event focuses on how intelligent systems interface with infrastructure, how they earn and distribute value, and how to preserve trust, privacy, and ownership as AI scales.
Join researchers, builders, and industry leaders for talks, demos, and hands-on sessions for a great two-day lineup.
Our Premier Partner: Roam
Roam is the virtual workspace our team relies on to stay connected across time zones. It makes collaboration feel natural with shared spaces, private rooms, and built-in AI tools.
Roam’s focus on human-centered collaboration is why they’re our Premier Partner, supporting our mission to connect the builders and leaders shaping the future of AI.
Experience Roam yourself with a free 14-day trial!
➡️ Before You Go
Partner With Us
Launching a new product or hosting an event? Put your work in front of our global audience of builders, founders, and operators — we feature select products and announcements that offer real value to our readers.
👉 To be featured or sponsor a placement, reach out to our team.
The AI Collective is a community of volunteers, made for volunteers. All proceeds directly fund future initiatives that benefit this community.
Stay Connected
💬 Slack: AI Collective
🧑💼 LinkedIn: The AI Collective
𝕏 Twitter / X: @_AI_Collective
Get Involved
About the Authors
About Noah Frank
Noah is a researcher, innovation strategist, and ex-founder thinking and writing about the future of AI. His work and body of research focus on aligning governance strategies to anticipate transformative change before it happens.
About AJ Green
AJ Green is a founder, writer, VC scout, chairman, and respected community leader in the AI and startup space. A former athlete turned tech entrepreneur, AJ is on a mission to make AI the great equalizer scaling startups, connecting ecosystems, and turning disruption into opportunity.
About Joy Dong
Joy is a news editor, writer, and entrepreneur at the forefront of the emerging tech landscape. A former educator turned media strategist, she demystifies complex systems to make AI and blockchain accessible for all. Joy is on a mission to explore how decentralized technology and artificial intelligence can be leveraged to build a more innovative and transparent future.