LLM Cost Calculator

Calculate LLM inference cost by task type, model tier, and volume. Compare API vs self-hosted pricing.

Task Type

Model Tier

Daily Volume

Avg Input Tokens

Avg Output Tokens

Response Cache Hit Rate: 20%

0% (no cache)25% (typical)50% (high)

Cost / Request

$0.130

Avg QPS

Peak: 36

Tokens / Day

2.5B

Daily Cost (API)

$104.0K

After 20% cache

API Pricing

Daily$104.0K

Monthly$3.12M

Yearly$37.44M

Self-Hosted (est.)

Daily$8.3K

Monthly$249.6K

Monthly Savings$2.87M

Estimates based on typical API pricing as of March 2026. Self-hosted estimates assume 8-12x cost reduction at scale (amortized GPU fleet). Actual costs vary by provider, volume discounts, and infrastructure choices.