AI Prompt Cost Calculator
FREEDeveloperInput Tokens
49
Output Tokens
500
Total Tokens
549
How many tokens the AI generates in response
For monthly cost projection
GPT-4o Mini
$0.0003 per call / $0.922 per month
99% cheaper than the most expensive option
| # | Model | Input $/1M | Output $/1M | Per Call | Monthly |
|---|---|---|---|---|---|
| 1 | ECO GPT-4o MiniCHEAPEST OpenAI | $0.15 | $0.6 | $0.0003 | $0.922 |
| 2 | ECO Gemini 3.1 Flash-Lite | $0.25 | $1.5 | $0.0008 | $2.29 |
| 3 | ECO Gemini 3.1 Flash | $0.75 | $3 | $0.0015 | $4.61 |
| 4 | ECO Claude Haiku 4.5 Anthropic | $1 | $5 | $0.0025 | $7.65 |
| 5 | MID GPT-4o OpenAI | $2.5 | $10 | $0.0051 | $15.37 |
| 6 | MID Gemini 3.1 Pro | $2 | $12 | $0.0061 | $18.29 |
| 7 | MID Claude Sonnet 4.6 Anthropic | $3 | $15 | $0.0076 | $22.94 |
| 8 | PRO GPT-5 OpenAI | $15 | $60 | $0.031 | $92.20 |
| 9 | PRO Claude Opus 4.6 Anthropic | $15 | $75 | $0.038 | $114.7 |
Monthly Cost Comparison (100 calls/day)
Highest intelligence, complex reasoning
Avg. monthly cost
$103.45
Strong performance, reasonable cost
Avg. monthly cost
$18.87
Fast and affordable for simple tasks
Avg. monthly cost
$3.87
AI API Pricing Reference (April 2026)
Per 1M tokens. Prices from official provider pricing pages.
| Model | Provider | Input $/1M | Output $/1M | Output/Input Ratio |
|---|---|---|---|---|
| Claude Opus 4.6 | Anthropic | $15 | $75 | 5.0x |
| GPT-5 | OpenAI | $15 | $60 | 4.0x |
| Claude Sonnet 4.6 | Anthropic | $3 | $15 | 5.0x |
| GPT-4o | OpenAI | $2.5 | $10 | 4.0x |
| Gemini 3.1 Pro | $2 | $12 | 6.0x | |
| Claude Haiku 4.5 | Anthropic | $1 | $5 | 5.0x |
| Gemini 3.1 Flash | $0.75 | $3 | 4.0x | |
| Gemini 3.1 Flash-Lite | $0.25 | $1.5 | 6.0x | |
| GPT-4o Mini | OpenAI | $0.15 | $0.6 | 4.0x |
About AI Prompt Cost Calculator
AI API pricing varies significantly across providers and model tiers. Anthropic, OpenAI, and Google all charge separately for input tokens (prompt + context) and output tokens (generated response), with output typically priced 3-5x higher than input. Token count estimation — converting raw text to an approximate token count before sending an API request — is essential for budgeting pipelines that process large volumes of text. Understanding token costs across 9 major models enables informed model selection for different task tiers.
How It Works
Token count estimation uses a hybrid approach: character-based estimation (approximately 4 characters per token for English text) averaged with word-based estimation (approximately 1.33 tokens per English word). The two estimates are averaged to reduce the directional bias of each method individually. For English text, this approximation is typically within 5-10% of the actual tokenizer output.
The cost calculation multiplies the estimated input token count by the model's input price per million tokens, adds the output token estimate multiplied by the output price per million tokens, and multiplies by the daily call volume to project monthly spend (using 30 days as the month length).
Model pricing is hardcoded from official pricing pages as of April 2026. The comparison table sorts models by monthly cost, allowing quick identification of the cheapest model that meets quality requirements for a given task. Context window limits are also shown to flag models that cannot accommodate the estimated prompt size.
Who Is This For
A backend engineer pastes a proposed system prompt into the calculator to estimate the monthly API cost of a customer support chatbot handling 10,000 conversations per day before committing to a model.
A startup CTO compares the monthly cost of Claude Sonnet 4.6 versus GPT-4o for a document summarization pipeline processing 500,000 pages per month, to determine which model fits the unit economics.
A developer debugging an expensive pipeline identifies that a 2,000-token system prompt repeated on every call accounts for 40% of total cost, motivating a switch to prompt caching.
A product manager evaluates whether upgrading from Gemini Flash-Lite to Gemini Pro is financially justified for a specific workflow, using the calculator to quantify the monthly cost difference.
How to Use
Paste your prompt or system message into the text area to estimate input tokens
Set the expected output length using the slider (100-4000 tokens)
Set your expected API calls per day for monthly cost projection
Compare costs across 9 major AI models sorted by price, tier, or monthly spend
Frequently Asked Questions
More Free Tools
View allRun your projects on Claude Code?
We packaged the .claude config that runs this site — 26 specialist agents, 14 workflow skills, and 6 rule files from 31 real production incidents. From $9.
Get the Claude Code Production Pack — $29