Anthropic x SpaceX

Anthropic is partnering with SpaceX to serve their models, which means higher usage limits for users.

Starting today:

Claude Code 5-hour rate limit: 2x higher

Peak-hour limit reduction: removed

API rate limits: raised

This could genuinely change the AI coding wars.

Anthropic’s biggest weakness was never model quality. It was the rate limit.

Compared to Codex and Gemini, Claude often felt too restricted. You try a few things, hit the limit, and suddenly you’re forced to switch tools.

But with more compute, a few things change:

People can experiment more without constantly worrying about limits.

Claude becomes easier to adopt as a daily coding tool.

We may see fewer sudden “nerfs” when demand spikes.

A lot of people moved to Codex recently because the rate limits were just more generous.

But if Claude Code can now give both strong model quality and higher usage limits, the competition gets way more interesting.

Which one are you using more now, Claude Code or Codex?