Five plans, three models, one decision. An opinionated map of what each tier of Claude is actually good for — with real use cases, performance trade-offs, and a calculator that tells you the honest answer.
Filter by audience. Click any card to highlight it; scroll down to see which one the calculator recommends for your workload. Prices below Anthropic's public list rates as of April 2026, annual billing where applicable.
All three current-generation models ship inside paid Claude plans. The differences are real but not absolute: Haiku handles most daily prompts fine, Sonnet is the workhorse, and Opus earns its keep on the hard stuff. Relative performance bars are calibrated from Anthropic's public benchmark positioning.
Pick a task on the left. The right panel shows the prompt, the recommended model, and a frank comparison of how each model would handle it — quality, speed, and whether the cost delta is justified.
A rough model, because Anthropic describes usage in relative terms (Max is 5x or 20x Pro, etc.) rather than exact message counts. Adjust your daily volume and task mix; the calculator estimates monthly cost per plan and flags the best fit. Treat the output as directional, not contractual.