A pivotal shift in AI monetization unfolded this week as Anthropic overhauled its subscription structure, while OpenAI launched a competitive promotion, igniting a full-scale pricing battle ahead of the Artificial Superintelligence (ASI) race. The moves target developers and enterprise users, reshaping how AI usage is billed and setting the stage for a fierce ecosystem showdown.
Anthropic’s Subscription Overhaul: From Unlimited to Capped Credits
Effective June 15, 2026, Anthropic is splitting its Claude subscription model into two separate pools, decoupling automated tool usage from interactive access. Previously, all activities—including Agent SDK calls, claude-p commands, and third-party integrations like OpenClaw and Hermes—shared a single subscription quota. Under the new policy, automated and programmatic usage will be stripped from base subscriptions and charged via fixed monthly API credits.
The credit allocations are tiered by plan:
- Pro: $20 monthly credit
- Max: $100 monthly credit
- Max 20x: $200 monthly credit
Interactive usage (Claude Code chat, claude.ai conversations) remains covered by unchanged subscription limits. Credits must be manually claimed via email notifications sent June 8, reset monthly with no rollover. To mitigate backlash, Anthropic doubled Claude Code’s 5-hour rolling limit last week and added a 50% boost to weekly quotas, effective through July 14 for all paid users.
OpenAI’s Counter: Free Codex Migration for Enterprises
Seizing the opportunity, OpenAI rolled out an aggressive response this week, offering two months of free Codex access for enterprise users migrating from Claude. The promotion includes a 30-day application window and a one-click migration tool that transfers Claude Code prompts, skills, and MCP configurations seamlessly. Unlike Anthropic’s split model, OpenAI’s Codex subscription unifies interactive and automated usage with no tiered restrictions.
Why Anthropic Made the Change
The restructuring addresses a critical flaw in unlimited subscriptions: unsustainable automated usage. Agent SDK and scripted calls run 24/7, consuming massive tokens with minimal training value. With its recent surge in compute capacity—including 220,000 GPUs from SpaceX, a $40 billion investment from Google, and a $25 billion capacity deal with Amazon—Anthropic faces exponential cost risks as its user base expands. The company explicitly states production-grade automation belongs to pay-as-you-go API plans, not fixed subscriptions.
The Cost of the Shift for Users
The credit overhaul drastically reduces effective limits for power users. A Max 20x plan with a $200 credit previously received $2,000–$5,000 worth of prorated automated usage. Now, the same $200 only covers API-priced calls. Pro users face even starker limits: $20 buys roughly 6–7 million input tokens or 1 million output tokens, exhausted quickly in intensive agent workflows. Third-party tool developers, once labeled "freeloaders" by Anthropic, now gain official recognition but rely solely on the meager monthly credits.
Misaligned Capacity vs. Developer Needs
Anthropic’s GPU surplus enables Claude Code’s quota hikes, but the credit caps alienate power developers. Interactive users benefit from expanded limits, while automation-focused developers bear the brunt. OpenAI exploited this divide, offering unrestricted usage and a frictionless migration path—critical for enterprises building long-term automated pipelines. The two-month free window lets businesses lock in workflows, raising switching costs permanently.
AI Subscription Trends & ASI Race Dynamics
The "all-you-can-eat" subscription era is ending. As agents and automation scale, pay-as-you-go pricing will dominate, mirroring cloud computing’s evolution. The battle for developer loyalty will define the ASI race, with two clear coalitions emerging:
- OpenAI + Microsoft + Stargate: Unified, unrestricted enterprise plans
- Anthropic + Google + Amazon + SpaceX: Tiered, capacity-controlled subscriptions
Key Milestones to Watch
Three events will shape the ecosystem’s balance:
- Mid-May: Musk vs. Altman legal ruling
- June: Anthropic’s IPO pricing
- July 14: Expiration of temporary Claude Code limits
Conclusion
Anthropic’s pricing fix and OpenAI’s counter mark the first major skirmish in the ASI race. While Anthropic safeguards profitability amid soaring compute costs, it risks ceding developer mindshare. OpenAI’s promotion capitalizes on user frustration, prioritizing ecosystem growth over short-term revenue. As AI capabilities converge, infrastructure strength and developer loyalty will determine the ultimate winner—making this pricing war a pivotal moment in AI’s commercial evolution.
