Mintlify just killed seat pricing in one sentence

In this edition:

• This week: Seat-based SaaS is dying, Anthropic split agent billing from chat subscriptions, and Anthropic passed OpenAI in enterprise spend for the first time

• Under the radar: OpenAI launched a $4B+ services arm, acquired 150 deployment engineers, and signed McKinsey, Bain, and Capgemini as partners

• What's on the calendar: Anthropic's June 15 billing change, the Musk v. Altman verdict, and WWDC

THE WEEK IN AI

THE WEEK IN ONE SENTENCE

Seat-based pricing dying, Anthropic splitting agent billing from chat, and Anthropic passing OpenAI in enterprise spend all point to the same thing: the unit of AI value is now the work the agent actually does.

THREE SIGNALS

01 • Pricing

Mintlify dropped seat-based pricing

Mintlify, the documentation platform used by Anthropic, OpenAI, Cursor, and most of the developer-tools market, killed seat-based pricing on Thursday. The announcement carried a line worth reading twice: "seat-based SaaS has no future where agents are the primary users."

Mintlify's documentation is read about as often by AI agents as by humans now. When a coding agent looks something up to write a pull request, the read is identical to a developer's read for billing purposes. Mintlify saw that pattern in their own usage data and stopped pretending the seat was the right unit.

The operator read: the next time your software vendor proposes a renewal, expect a different shape on the page. Consumption pricing is moving from rare to standard in any tool an agent might call.

AI READY PRO · FREE UNTIL FRIDAY

I will get you AI trained in 30 days for free

After several months of running it quietly with top AI operators and teams, I’m excited to launch AI Ready Pro today to our newsletter subscribers like you.

It’s a 30-day personalized AI training program that is personalized to you and where you are in your AI journey. It’s the culmination of 100’s of hours spent teaching AI to top Fortune 500 AI teams and operators (and working with Mark Cuban).

Each day, you receive a 10-15 minute exercise to complete to improve your AI skills.

This isn’t vague AI theory or a way to pitch you a tool. It’s 30 days of learning to use AI in your real work. So you come out the other side with the skills to use AI to actually improve you and your teams output.

Until this Friday (May 15), it’s free for newsletter readers who complete the extended assessment. This assessment will help us personalize the learning experience to you.

If you lead a team, AI Ready Team is open today too, also for readers first. It’s the same engine, but to train your whole organization. You’ll get a detailed view of:

• Your team’s AI readiness level + detailed strengths & weaknesses

• Ranked list of AI automation opportunities

• Get tactical advice on how to advance AI efforts in your org

Leading a team? Take the Team assessment instead.

02 • Adoption

Anthropic split Agent SDK billing from chat subscriptions

Starting June 15, programmatic Claude usage through the Agent SDK gets its own credit pool. It will be separate from the $200/month Max chat subscription that many developers had been quietly using for agent work.

The plain reading is that Anthropic noticed customers running production agents on a consumer subscription and decided that math no longer worked for either side. The Agent SDK has a different cost structure: longer context, more concurrent calls, less idle time. A human chat user costs one thing. An agent loop costs a different thing.

What this means for an operator: the agent budget is not the same as the AI tools budget. If your team has been folding Claude into a flat per-user line item in finance, the June change forces those into two lines. The conversation with your CFO becomes about what the agent is doing, not how many people are logging in.

03 • Market

Anthropic passed OpenAI in enterprise adoption for the first time

The Ramp AI Index released Thursday shows Anthropic at 34.4 percent of enterprise AI spend tracked through Ramp corporate cards, versus OpenAI at 32.3 percent. Anthropic quadrupled its share over the past year while OpenAI grew 0.3 percentage points.

This is the first time the order has flipped on this index. Two things to read into it without overreading. First, Ramp's data captures small-to-mid-market spend better than Fortune 500 contracts. The headline is more about who is winning the broad middle than who is winning the largest accounts. Second, Anthropic's lead lines up with where Claude Code has been deployed: coding agents that operators bought on consumption-priced credits, not per-seat ChatGPT Enterprise rollouts.

The two pricing signals above and the share shift in this one fit together. The buyer reallocated toward the vendor whose pricing fit how the tools are actually used.

UNDER THE RADAR

The story most people pattern-matched as "OpenAI launched a services arm" is worth a closer look. On Tuesday, OpenAI launched the OpenAI Deployment Company, acquired Tomoro (about 150 Forward Deployed Engineers), and announced 19 institutional partners. The partner list includes TPG, Advent, Bain Capital, Brookfield, Goldman Sachs, SoftBank, Warburg Pincus, plus Bain & Company, Capgemini, and McKinsey & Company. Initial commitment: more than $4 billion.

Read that partner list slowly. The PE and growth funds are there because their portfolio companies are the buyers. McKinsey, Bain, and Capgemini are there because they bring the implementation channel. The new entity is majority-owned by OpenAI, which means OpenAI now controls the channel as well as the model.

I've been spending the last two years inside operator-AI implementations, and one pattern keeps repeating. The constraint is rarely the model. It is whether anyone inside the company can redesign workflows around it without ripping things in half. OpenAI just bought 150 people who do that work and signed up the world's largest consulting firms to do it at scale. The pricing changes above set the unit. This sets the distribution.

The operator question worth sitting with this weekend: who is going to do this work inside your company, and what do you want them to look like? An internal team, a partner, a Forward Deployed Engineer on loan from OpenAI, or some mix. The answer will not be the same in 12 months as it is today.

QUOTE OF THE WEEK

❝

“Seat-based SaaS has no future where agents are the primary users”

Mintlify, May 14, 2026

SPONSORED BY CLUTCH

Hire secure AI teammates that work 24/7.

Hire pre-built AI teammates. Give your engineers and operators a platform to ship their own AI apps. Stop losing sleep about what is running where.

Clutch is the platform behind both: pre-built agents for the workflows your ops team should automate first, plus the integration plane your team's vibe-coded apps and Claude Code projects plug into. One platform. Real production. Visible and safe by default.

Built for ops, engineering, and security teams that are tired of the shadow-AI surface area inside their own company.

WHAT’S ON THE CALENDAR

• June 15: Anthropic's Agent SDK billing change takes effect. Worth pulling your team's Anthropic invoice before and after to see what the split looks like in practice.

• Musk v. Altman: Jury began deliberating Thursday after closing arguments. A verdict in the next week or two would reshape how OpenAI talks about its founding story for the rest of the year.

• OpenAI weekly cadence: OpenAI confirmed it is moving to a stable Thursday release schedule. Expect smaller increments, more frequently.

• WWDC: Apple's developer conference is mid-June. Watch whether AI glasses, on-device model strategy, or anything resembling an OpenAI competitor surfaces.

P.S.

P.S. If your team is renegotiating an AI tool contract in the next 90 days and you want a second set of eyes on the pricing model, hit reply. I read every reply, and the Mintlify line is going to keep showing up in different versions for the rest of the year.

Forward this to one person who is about to sign a renewal.

Have a good weekend,
Haroon