News
Research Report
Mind The Agentic Action Gap: Stop Losing Money With AI Agents
Key Takeaways
- Anthropic’s SpaceX deal gives it access to more than 300 megawatts of new compute capacity.
- Claude Code users on Pro, Max, Team and Enterprise plans will get doubled five-hour rate limits.
- Anthropic is removing peak-hour limit reductions for Claude Code users on Pro and Max plans.
- The deal shows how compute access is becoming a direct driver of AI product availability.
Anthropic announced a new partnership with SpaceX to use the full compute capacity at SpaceX's Colossus 1 data center. According to Anthropic, the deal provides access to more than 300 megawatts of new capacity — over 220,000 NVIDIA GPUs — within the month.
The added infrastructure enabled Anthropic to immediately raise usage limits across Claude Code and the Claude API. The company has also expressed interest in partnering with SpaceX to develop multiple gigawatts of orbital data center capacity.
The SpaceX deal joins a series of compute agreements that now includes an up to 5 gigawatt (GW) deal with Amazon (nearly 1 GW by end of 2026), a 5 GW agreement with Google and Broadcom (online in 2027), a $30 billion Azure partnership with Microsoft and NVIDIA and a $50 billion US infrastructure investment with Fluidstack.
Table of Contents
- Claude Gets More Room to Run
- How Compute Growth Reshapes AI Throughput
- Claude’s Growth Spurt Comes With New Scrutiny
Claude Gets More Room to Run
Anthropic said the SpaceX partnership will translate directly into higher limits for Claude Code and the Claude API, with the changes taking effect May 6, 2026. The company framed the move as an immediate capacity upgrade for its most active users.
For Claude Code users, Anthropic is doubling five-hour rate limits for Pro, Max, Team and seat-based Enterprise plans. That means developers using Claude Code for longer coding sessions should have more room to work before hitting usage ceilings.
The company is also removing peak-hour limit reductions for Claude Code users on Pro and Max accounts. In practical terms, that should make Claude Code access more predictable during high-demand periods, when throttling or reduced availability can interrupt developer workflows.
Anthropic is also raising API rate limits for its Claude Opus models:
| Tier | Maximum Input Tokens Per Minute | Maximum Output Tokens Per Minute |
|---|---|---|
| 1 | 30,000 → 500,000 | 8,000 → 80,000 |
| 2 | 450,000 → 2,000,000 | 90,000 → 200,000 |
| 3 | 800,000 → 5,000,000 | 160,000 → 400,000 |
| 4 | 2,000,000 → 10,000,000 | 400,000 → 800,000 |
How Compute Growth Reshapes AI Throughput
Architecture choices behind large-scale capacity expansion shape API limits, latency and developer throughput as much as the capital funding them.
Major AI labs have moved away from single-provider infrastructure. Anthropic's multi-billion dollar TPU deal with Google, announced late 2025, provides access to up to one million Tensor Processing Units (TPUs), projected to bring more than one gigawatt online by 2026.
Anthropic distributes workloads across Google's TPUs, Amazon's Trainium chips and NVIDIA GPUs, each handling specialized tasks including training, AI inference and research.
Not all compute expansion is hyperscale. In enterprise-dense markets like New York City, colocation demand reflects latency requirements and proximity to financial customers — not large-scale model training. Inference workloads can tolerate smaller footprints, making urban multi-story facilities viable.
Claude’s Growth Spurt Comes With New Scrutiny
Anthropic closed three consecutive funding rounds in under a year:
- A $3.5 billion Series E in March 2025 at a $61.5 billion valuation
- A $13 billion Series F led by ICONIQ in September 2025
- A $30 billion Series G in February 2026 at a $380 billion valuation.
Webinar
May
12
How Marketing Teams Can Take Control of the Web Experience
See how Veo and Kvalifik put marketing in the driver's seat, shipping faster and growing qualified bookings by 150%.
Register
Webinar
May
14
Two Audiences, One WordPress. Is Your Site Ready?
The practical steps to making your WordPress portfolio ready for both human visitors and AI agents.
Register
Webinar
May
19
From Content Sprawl to Competitive Advantage: A Knowledge Operations Roadmap
How automation, AI and a centralized content platform turn knowledge fragmentation into a scalable competitive advantage.
Register
Webinar
May
20
From AI Investment to CX Results: What Enterprise Leaders Need to Know
Move beyond experiments. See how top enterprises scale AI for CX results.
Register
Webinar
May
20
AI With Intent: How CX Leaders Decide Where Automation Belongs (And Where It Doesn't)
Senior CX leaders from REI & Hyatt reveal their framework for using AI to boost retention.
Register
Webinar
Jun
16
From Legacy to Launch-Ready: How Gainbridge Made Its Website a Marketing-Led Growth Engine
Join in to learn how a D2C annuity brand gave marketing full website ownership — without slowing down or risking compliance.
Register
Revenue scaled from roughly $1 billion in annualized run rate in January 2025 to approximately $20 billion by early March 2026, per Bloomberg — with Claude Code alone surpassing $2.5 billion in run-rate revenue by February 2026.
Product launches have kept pace. Anthropic shipped the Claude Code Slack integration in December of last year, the Cowork desktop automation agent in January and saw Apple integrate the Claude Agent SDK into Xcode 26.3 in February.
Not all of the momentum has been positive, however. In early March 2026, three US cabinet agencies — State, Treasury and Health & Human Services — directed staff to stop using Claude following a dispute over military-use guardrails. CEO Dario Amodei publicly criticized OpenAI after the company pursued the same Pentagon contract without equivalent safety measures.