Version 3

On May 6, 2026, Anthropic held its 'Code w/ Claude 2026' developer conference, with Simon Willison attending and live-blogging the morning keynote sessions.[1] The event was paired with a same-day announcement of major infrastructure and product changes: Anthropic signed an agreement with SpaceX to access all compute capacity at the Colossus 1 data center — over 300 megawatts and 220,000+ NVIDIA GPUs — with access expected within the month.[2] Alongside the compute deal, Anthropic doubled Claude Code's five-hour rate limits for Pro, Max, Team, and seat-based Enterprise plans, and removed the peak-hours rate-limit reduction for Pro and Max subscribers, framing these as a direct consequence of the new capacity.[2]

The SpaceX partnership is embedded in a broader infrastructure buildout: Anthropic disclosed a multi-provider compute portfolio exceeding 10 GW of future capacity spanning Amazon, Google/Broadcom, Microsoft/NVIDIA, and Fluidstack, and expressed interest in working with SpaceX to develop 'multiple gigawatts of orbital AI compute capacity.'[2] Anthropic explicitly tied its partner selection to geopolitical criteria, stating it would only add capacity in democratic countries with supportive legal frameworks and secure hardware supply chains.[2] The scale of these ambitions stands in contrast to a practical supply complaint that surfaced days earlier: on May 1, users reported that Claude Opus 4.7 quotas on AWS Bedrock had abruptly dropped to 0 tokens per minute.[3]

By May 6–7, the story had migrated from developer forums and Anthropic's own channels into mainstream tech press. Ars Technica reported straightforwardly on the rate-limit increases and their credited cause.[4] Wired's coverage adopted a sharper editorial tone — its headline 'Anthropic Gets in Bed with SpaceX' carries connotations of an uncomfortable or ethically complex entanglement, signaling that at least some outlets are framing the deal with skepticism rather than celebration.[5] This framing shift is notable given Anthropic's safety-first brand positioning and its explicit geopolitical partner criteria; a partnership with a company controlled by Elon Musk invites scrutiny of those stated values.

In the weeks leading up to the event, the developer community generated a stream of real-world Claude Code adoption stories: Fiberplane published a detailed account of how it uses Claude Code and agents internally[6], one user described migrating 16 websites between Linode servers in a single day with Claude Code's help[7], and a technical post explored pro-grade Java development in Claude Code using LSP integration beyond simple grep.[8] Competitive context also sharpened: OpenAI was reported to be targeting Anthropic directly with $100/month ChatGPT Pro subscriptions in early April.[9] Some developers on Hacker News were simultaneously asking whether any open-source model and harness had reached Claude Sonnet or Opus performance levels[10], suggesting that even as Anthropic builds out capacity, a segment of users is watching for credible alternatives.