Input vs output token spend
Output tokens usually cost several times more than input, so answer length is what really drives your bill.
Home / Dev & AI / AI & LLM Cost / Claude API Cost Calculator
AI & LLM Cost
Estimate your Anthropic bill fast. Pick a Claude model, enter your request volume and token sizes, and see the monthly cost with a comparison across Opus, Sonnet and Haiku for 2026.
Estimated monthly cost
$0
Input vs output token spend
Output tokens usually cost several times more than input, so answer length is what really drives your bill.
The same workload priced across AI models
Your monthly token volume run through popular 2026 models, cheapest first.
Anthropic bills the Claude API per token, and the three tiers sit far apart on price. Claude Haiku is the fast, low cost option, Claude Sonnet is the balanced workhorse most apps run on, and Claude Opus is the premium model for the hardest reasoning. Because Opus costs many times more than Sonnet, matching the tier to the task is the key budgeting decision.
This Claude API cost calculator prices your workload on the model you choose and compares it across all three tiers, so you can see exactly how much a step up to Opus or down to Haiku changes your monthly bill.
On Claude Sonnet 4.5, 100,000 requests a month at 1,000 input and 500 output tokens cost 100M × $3 + 50M × $15 per million = $300 + $750 = about $1,050 a month. The same on Claude Haiku 4.5 is roughly $350, while Opus 4.1 jumps to about $5,250.
| Model | Input / 1M | Output / 1M |
|---|---|---|
| Claude Opus 4.1 | $15.00 | $75.00 |
| Claude Sonnet 4.5 | $3.00 | $15.00 |
| Claude Haiku 4.5 | $1.00 | $5.00 |
July 2026 list estimates. Anthropic offers prompt caching and batch processing that can cut costs substantially on repeated context and non-urgent jobs.
Claude Sonnet 4.5 is about $3 per million input tokens and $15 output. Haiku is cheaper and Opus is the premium tier. Enter your usage above for a tailored figure.
Sonnet suits most production work. Use Haiku for high volume, latency sensitive tasks, and reserve Opus for the most demanding reasoning where quality justifies the price.
Yes. Caching a stable system prompt or long context avoids re-billing those tokens at full rate on every call, which is a large saving for chat and agents.
It depends on the tier. Sonnet is priced above GPT-5 on input but competitive overall, while Haiku undercuts most premium models. Compare on your own tokens to be sure.
They are July 2026 list estimates. Confirm current rates on the Anthropic pricing page before budgeting.