LLM Token Cost Calculator
Enter input and output token counts to instantly compare costs across major LLM APIs. All calculations run locally. See Pricing Comparison for primary sources.
💡 Rough guide: English ≈ 1.3 tokens/word. Japanese ≈ 1–1.3 tokens/char. Varies by content.
Default reference rate. Adjust as needed.
Comparison Results (sorted by total cost)
| Model | Provider | Input Price | Output Price | Total (USD) | Approx. JPY |
|---|
※ Standard (non-cached) rates only. Batch API & prompt cache discounts not included. Claude Fable 5.1: $10/$50 (cache read $0.25, released 2026-09-01). Claude Sonnet 5: $2/$10 (now permanent). GPT-5.6 Sol: promo rate (thru 11/21). GLM-5.3-Flash: list price ($0.15/$0.50; promo ended 9/9). DeepSeek: V4.1 Flash is now the workhorse ($0.15/$0.60 off-peak, $0.30/$1.20 peak) — both peak/off-peak listed (peak = weekdays only, 01-04 & 06-10 UTC, weekends all off-peak). Requests to V4 Flash are served by V4.1 Flash at the Flash price. V4 Pro keeps running after 2026-09-14 (the 9/14 routing notice was withdrawn) and its rates remain current. GPT-5.6 Terra/Luna: post-July-30 cuts. Always check official pages. See Pricing Comparison.