LLM API cost calculator
| Model | Per request | Per day | Per month |
|---|---|---|---|
GLM-5.3-Flashglm-5.3-flash30% off | $0.000384 | $0.3838 | |
Grok 4.7grok-4.790% off | $0.000696 | $0.6964 | |
DeepSeek-V4.1-Flashdeepseek-v4-1-flash30% off | $0.000837 | $0.8374 | |
GLM-5.3glm-5.330% off | $0.003489 | $3.49 | |
Kimi K3kimi-k330% off | $0.009421 | $9.42 |
Your price on cheapaimodel in USD; struck-through is the list price; a month is 30 days.
How the cost is calculated
Formula
per request = (uncached input × input price + cached input × cache-read price + output × output price) ÷ 1,000,000 per day = per request × requests per day per month = per day × 30
Worked example: GLM-5.3-Flash
The first row of the table, with the numbers you entered. Prices are per million tokens.
(2,000 × $0.104676 + 0 × $0.020935 + 500 × $0.348919) ÷ 1,000,000 = $0.0003838 $0.0003838 × 1,000 = $0.3838 $0.3838 × 30 = $11.51
What to enter
Input tokens are everything you send in one request: the system prompt, the conversation so far and the new message. Output tokens are what the model writes back. Cached input is the share of input that the provider reads from its prompt cache, billed at the lower cache-read price.
Which prices
The table uses the live prices you pay on cheapaimodel. A struck-through monthly amount is the same usage at the vendor's list price. If a model has no cache-read price, cached tokens are counted at its input price, so the estimate errs on the high side. Models billed per request use their price per call.
Finding your numbers
Each API response has a usage field with prompt and completion tokens. Average them over a day of traffic to fill in the inputs. To size one prompt before you send it, paste it into the token counter.