IMPLICATOR
Anthropic widened its lead in Implicator's weekly LLM Meter to 89 from 87, after the company announced a roughly $1.5 billion Wall Street joint venture on May 4, ten financial-services agent templates with Excel, PowerPoint, and Word generally available and Outlook in beta on May 5, and a SpaceX deal on May 6 covering all the AI capacity at the Colossus 1 data center in Memphis. Google's Gemini rose to 87 from 86 after the Pentagon cleared it for Impact Level 6 and Impact Level 7 classified networks on May 1 alongside OpenAI, Microsoft, AWS, Nvidia, SpaceX, and Reflection. OpenAI's ChatGPT moved to 83 from 82 on a parallel $10 billion Deployment Company joint venture announced the same day as Anthropic's.
Key Takeaways
- Anthropic moved to 89 from 87 after a $1.5 billion Wall Street venture, ten finance agents with Microsoft 365 GA, and a SpaceX deal locking up Colossus 1.
- Google's Gemini rose to 87 from 86 after Pentagon clearance for IL6/IL7 classified networks alongside six other AI vendors on May 1.
- OpenAI's ChatGPT moved to 83 from 82 on a parallel $10 billion Deployment Company joint venture announced the same day as Anthropic's.
- Grok fell to 33 and DeepSeek slipped to 14 on Pentagon exclusion and a Canadian penetration test, respectively.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
Wall Street venture targets the consulting budget
Blackstone, Hellman & Friedman, and Goldman Sachs partnered with Anthropic on the May 4 venture, Anthropic confirmed in its announcement. Anthropic, Blackstone, and Hellman & Friedman each committed roughly $300 million and Goldman Sachs added $150 million, Bloomberg reported. Apollo Global Management, General Atlantic, Leonard Green, GIC, and Sequoia Capital backed the consortium. The mandate places Anthropic engineers inside portfolio companies of the participating private-equity firms and puts Claude in direct competition with McKinsey, BCG, Bain, and Accenture for corporate AI transformation work.
The financial-services briefing the next day extended the distribution thesis. Anthropic shipped ten Claude agent templates for pitchbook creation, credit memos, KYC reviews, financial modeling, and month-end close. Microsoft 365 integration moved to general availability for Excel, PowerPoint, and Word, with Claude for Outlook in beta. Moody's embedded its full platform into Claude as a native app covering credit ratings and risk data on more than 600 million companies, Fortune reported. Verisk, Third Bridge, Fiscal AI, Dun & Bradstreet, Experian, GLG, Guidepoint, and IBISWorld joined an existing data-partner roster of LSEG, S&P Capital IQ, Morningstar, and PitchBook.
SpaceX deal locks up Colossus 1 capacity
Anthropic took all the AI capacity at SpaceX's Colossus 1 data center in Memphis on May 6, gaining access to more than 300 megawatts and over 220,000 Nvidia GPUs within the month, Anthropic disclosed in a release. The accelerator mix spans H100, H200, and GB200 chips per Tom's Hardware. The company also stated interest in working with SpaceX on multi-gigawatt orbital compute capacity. The day after the deal, Anthropic doubled Claude Code limits, removed peak-hours throttling for Pro and Max subscribers, and raised Opus API rate limits.
The full enterprise LLM scoreboard, every Sunday
Strategic AI news from San Francisco. No hype, no "AI will change everything" throat clearing. Just what moved, who won, and why it matters. Daily at 6am PST.
No spam. Unsubscribe anytime.
OpenAI mirrors with $10B Deployment Company
OpenAI finalized The Deployment Company on the same day Anthropic announced its venture, anchoring a $10 billion vehicle with TPG, Brookfield Asset Management, Advent International, Bain Capital, and Goanna Capital, Bloomberg reported. The fund raised more than $4 billion across 19 investors, with a 17.5% guaranteed annual return over five years. OpenAI is contributing $500 million up front with an option to add up to $1.5 billion. GPT-5.5, released on April 23, hit 82.7% on Terminal-Bench 2.0 and saw API revenue grow more than two times faster than any prior OpenAI launch in its first week. The Pentagon clearance on May 1 restored federal-tier procurement parity Anthropic lost in February when the Department of Defense designated the company a supply-chain risk.
Grok and DeepSeek lose ground
Grok fell to 33 from 35 after the Pentagon classified-network roster excluded xAI despite SpaceX appearing on it, and after SpaceX rented all the Colossus 1 capacity to Anthropic in the same week. Coverage of the Musk-Altman trial continued, with Elon Musk's April 30 testimony that xAI "partly" used distillation on OpenAI models still in circulation. Mistral rose to 74 from 73 on the May 2 launch of Mistral Medium 3.5 and Vibe Remote Agents, with the system posting a 77.6% score on SWE-Bench Verified. DeepSeek slipped to 14 from 15 after Canada's Communications Security Establishment ran a penetration test that extracted system-level instructions and bypassed moderation filters via phonetic obfuscation specific to DeepSeek's tokenizer.
The next LLM Meter posts on May 17, as the Musk-Altman trial continues in California federal court with Sam Altman and Greg Brockman expected to testify.
Frequently Asked Questions
How is the LLM Meter scored?
Implicator scores six tracked models each week from an enterprise-buyer perspective, weighing compliance and security, model quality on professional tasks, reliability and SLAs, ecosystem maturity, vendor stability, and pricing. Scores range from 0 to 100 and publish weekly on Sundays.
Why did Anthropic gain ground this week?
Three concurrent moves shifted the score: a $1.5 billion Wall Street venture with Goldman Sachs, Blackstone, and Hellman & Friedman; ten financial-services agent templates with full Microsoft 365 integration on May 5; and a SpaceX deal on May 6 covering all the AI capacity at the Colossus 1 data center in Memphis.
Why was Grok marked down?
The Pentagon's May 1 classified-network roster cleared seven AI vendors for IL6 and IL7 deployments and excluded xAI even though SpaceX appeared on the list. SpaceX then rented all of Colossus 1's compute to Anthropic, a direct rival, in the same week.
What does the OpenAI Deployment Company do?
The Deployment Company is a $10 billion vehicle anchored by TPG, Brookfield, Advent International, Bain Capital, and Goanna Capital. It embeds OpenAI engineers inside portfolio companies in healthcare, logistics, manufacturing, and financial services, mirroring the distribution model Anthropic announced the same day.
What is the next major event for the meter?
The May 17 LLM Meter posts as the Musk-Altman trial enters a second week of testimony in California federal court, with Sam Altman and Greg Brockman expected on the stand. Training-data attestation and OpenAI's IPO posture are the most likely score movers.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
Marcus Schuler
Editor-in-Chief and founder of Implicator.ai. Former ARD correspondent and senior broadcast journalist with 10+ years covering tech. Writes daily briefings on policy and market developments. Based in San Francisco. E-mail: [email protected]
