Anthropic subscriptions estimated at roughly 5x OpenAI’s API-equivalent value for mid-tier models

SemiAnalysis says its usage-limit tests put Anthropic at roughly five times OpenAI’s API-equivalent value when comparing Opus 5.5 with GPT 6.1 Sol under an agentic workload. It says limits for GPT-6 Astra and Fable 5.1 are much closer, and cautions that a plan’s value varies by model and workload. Its tests also found that OpenAI halved token limits for newly purchased $200 plans.