AI Calculator Pro
Value

LLM API Cost Calculator

Estimate the cost of any LLM API call per request and per month.

Quick answer

For the “Single request” scenario, $0.00750 per request — ~$0.2250/month at 1 requests/day. Enter your own numbers below to recompute instantly; the full step-by-step math is shown under the worked example.

Results update automatically as you type.

Result
$0.00750 per request
~$225.00/month at 1,000 requests/day
Input cost
$0.00250
Output cost
$0.00500
Per request
$0.00750
Per day
$7.50
Per month
$225.00

Estimate what any large language model API will cost you. Enter your input and output token counts, pick a model, and see the cost per request, per day, and per month. Use it to budget an app, compare models, or sanity-check a provider's bill before you commit.

Not sure what to enter? Typical token ranges

Common engineering tasks: bug fixes, endpoints, tests, refactors. These are honest planning ranges for how many tokens an AI coding agent burns to build each item — the cost column uses GPT-4o’s current prices.

Work itemTypical input tokensTypical output tokensTypical cost
Small bug fix100,0006,000$0.3100
Complex bug fix400,00020,000$1.20
AI code review of a PR200,0008,000$0.5800
Unit test suite for a module380,00035,000$1.30
CRUD API endpoint300,00020,000$0.9500
Third-party API integration600,00035,000$1.85
Refactor a module750,00042,000$2.30
Data / schema migration520,00030,000$1.60
Technical documentation220,00025,000$0.8000

Ranges assume a whole agent session (many turns, context re-sent), not a single prompt. See how many tokens does X take for full low/typical/high ranges, or open the Cost to build X estimator to plug in your own model.

How this is calculated

Cost per request = (input tokens ÷ 1,000,000 × input price) + (output tokens ÷ 1,000,000 × output price). Input and output are priced separately, and output is usually 3–5× more expensive per token, so long generations dominate the bill.

We pull the input and output price for your chosen model from our live pricing data (refreshed daily) and multiply by your token counts. Daily and monthly figures scale the per-request cost by requests per day × 30. Prompt caching reduces the price of the cached share of input tokens; batch pricing swaps in the provider's discounted batch rates when enabled.

Is this a good result? What to do next

Judge the per-request number against your revenue per request, not in isolation. A $0.01 call is cheap for a paid feature but expensive if you run it free millions of times a day. The monthly figure is what matters for budgeting — multiply it out to a year and check it against the value the feature creates.

Typical planning ranges

Budget/mini models (e.g. GPT-4o mini, Gemini Flash)
~$0.15–$0.60 per 1M tokens
Mid-tier models (e.g. GPT-4o, Claude Sonnet)
~$3–$15 per 1M tokens
Frontier/reasoning models
~$15–$75 per 1M output tokens
Typical short request (1k in / 500 out, mid-tier)
~$0.005–$0.02

Ranges are typical planning figures to sanity-check your result, not authoritative benchmarks. Your numbers will vary with use case, volume, and vendor.

How to improve this number

  • Route simple requests to a cheaper/smaller model and reserve frontier models for hard ones.
  • Trim the system prompt and cap max output tokens — output is the expensive side.
  • Enable prompt caching for repeated context (system prompts, few-shot examples).
  • Use batch pricing for non-real-time jobs to get the discounted rate.

Common mistakes

  • Forgetting output tokens are billed at a higher rate than input.
  • Estimating with average tokens when a few large requests skew the real bill.
  • Ignoring retries, tool-call round-trips and moderation calls that add hidden requests.

When to use a different approach

If you resend a growing conversation history each turn, use the chatbot cost calculator, which models history growth. If you're comparing many models for one workload, use the model cost comparison.

Worked examples

Real scenarios you can reproduce by loading these presets in the calculator above:

Single request

Result
$0.00750 per request
~$0.2250/month at 1 requests/day
Input cost
$0.00250
Output cost
$0.00500
Per request
$0.00750
Per day
$0.00750
Per month
$0.2250
  • input: 1,000 / 1e6 x $2.5/M = $0.00250
  • output: 500 / 1e6 x $10/M = $0.00500
  • monthly = $0.00750 x 1 x 30 days

Small app (1k/day)

Result
$0.00750 per request
~$225.00/month at 1,000 requests/day
Input cost
$0.00250
Output cost
$0.00500
Per request
$0.00750
Per day
$7.50
Per month
$225.00
  • input: 1,000 / 1e6 x $2.5/M = $0.00250
  • output: 500 / 1e6 x $10/M = $0.00500
  • monthly = $0.00750 x 1,000 x 30 days

Scale (100k/day)

Result
$0.00750 per request
~$22,500.00/month at 100,000 requests/day
Input cost
$0.00250
Output cost
$0.00500
Per request
$0.00750
Per day
$750.00
Per month
$22,500.00
  • input: 1,000 / 1e6 x $2.5/M = $0.00250
  • output: 500 / 1e6 x $10/M = $0.00500
  • monthly = $0.00750 x 100,000 x 30 days

Sources & references

Frequently asked questions

How is LLM API cost calculated?+

Cost = (input tokens / 1,000,000 x input price) + (output tokens / 1,000,000 x output price). Providers price input and output tokens separately, and output is usually more expensive.

What are input vs output tokens?+

Input tokens are everything you send (system prompt, history, user message). Output tokens are what the model generates. You pay for both, at different rates.

How can I reduce cost?+

Use a smaller or cheaper model, shorten prompts, cap output length, enable prompt caching for repeated context, and use batch pricing for non-real-time jobs.

Related calculators

API Cost
AI Chatbot Cost Calculator

Estimate the monthly cost of running an AI chatbot at scale.

Comparisons
LLM Model Cost Comparison

Compare the cost of every model for the same workload.

Prompt Engineering
Context Window Planner

Check whether your prompt, history and context fit a model's window.

Answering a real question?

This calculator powers these problem-solving guides: