Claude’s pricing has changed more than once this year, which makes it worth understanding the structure rather than a specific dollar figure that might be stale by the time you read it.
There’s a free tier with usage limits, a Pro subscription for individuals with higher limits and access to the newest models, and API access billed per token, with rates varying by model. Check claude.ai directly for current numbers, not a cached figure from an article.
One detail worth knowing if you’re budgeting API usage: our coverage of Claude Sonnet 5’s pricing flags that a new tokenizer can count meaningfully more tokens for the same text than the previous one did. A rate that looks unchanged on paper can still mean a real cost increase. Re-benchmark your actual token usage after any model upgrade.
Prompt caching cuts costs for workflows that reuse the same context repeatedly. Using a smaller, cheaper model for high-volume, low-complexity tasks and saving the larger model for genuinely hard problems helps too. And check whether your use case is already covered by a Pro subscription before paying for separate API access.
Check current Claude pricing directly at claude.ai/pricing.
