Anthropic secretly A/B tests Claude limits: same $200 plan, ~20% less usage
RexDouglass · x · 2026-10-24
SemiAnalysis detected an A/B test on Claude Max limits: same $200/month, different quotas. Testing two Max accounts the same day on the same model, one got 19% more per 5-hour window (1,740M vs 2,076M tokens) but 20% LESS weekly budget (6,579M vs 5,266M) — monthly API-equivalent value dropping from $11,726 to $9,385. The bigger 5-hour window looks like an upgrade, but daily Claude Code users hit the weekly cap sooner, roughly $2,300/month less in API-equivalent usage. A Claude Code engineer told a user "our limits haven't changed" shortly before this surfaced. Anthropic hasn't confirmed; only some accounts are affected.
Related event: SemiAnalysis Finds Anthropic A/B Testing Claude Usage Limits(2 posts)→
More from Models
- LiveBench creator to evaluate Kimi K3.1, betting his benchmark can't be gamed — bindureddy · 2026-10-24
- Kimi K3.1 Pro tops SWE Marathon and agent benchmarks, fueling open-source 'last time they catch up' debate — cephaloform · 2026-10-24
- Kimi K3.1 Pro tops Humanity's Last Exam with tools at 70.0, ahead of Opus 5.5 — UltraRareAF · 2026-10-24
- Peter Gostev demos 3D generation with Kimi K3.1 Pro — arena · 2026-10-24
- Kimi K3.1 Pro hits #4 lab on Code Arena WebDev, 74% cheaper than Claude Opus 5 — arena · 2026-10-24
- Moonshot Kimi now ranks #4 lab on Code Arena WebDev leaderboard — arena · 2026-10-24