AI Prompt Cost Calculator
See how many tokens your prompt uses and what it costs across GPT-4o, Claude, Gemini, DeepSeek, Llama and 15+ models. Free. Instant. No signup.
How It Works
Paste Your Prompt
Drop in any text. We tokenize it in real-time using the same encoder OpenAI uses.
Set Your Volume
Set expected output length and daily request count. We calculate the rest.
Compare & Ship
See every model side-by-side. Find the cheapest option for your use case and ship with confidence.
Why This Exists
The difference between GPT-4o and GPT-4o Mini on a 1,000-token prompt is 16x. At 10,000 requests per day, that's the difference between $750/month and $45/month. Most teams don't check until the invoice hits.
AI model pricing is fragmented across dozens of provider pages, each with different formats, units, and gotchas. Some charge per character, some per token, some per "input unit." Context window limits, cached input discounts, batch pricing — it adds up to a mess.
PromptCost puts every model on the same page, in the same units, with the same math. Paste your actual prompt, set your expected volume, and see exactly what you'll pay. No signup, no API key, no sales call. Just the numbers.
Frequently Asked Questions
Tokens are the basic units AI models use to process text. A token can be a word, part of a word, or a single character. For English text, 1 token is roughly 4 characters or about 0.75 words. Every API call is priced based on the number of tokens processed — both the tokens you send (input) and the tokens the model generates (output).
We multiply your input tokens by the model's input price per token, plus the expected output tokens by the output price per token. That gives you the cost per request. Multiply by your daily request volume and by 30 to get your monthly estimate. All prices are per 1 million tokens, as listed on each provider's official pricing page.
GPT-4o costs $2.50 per million input tokens and $10.00 per million output tokens. A typical request with 500 input tokens and 500 output tokens costs about $0.00625 — or $6.25 per 1,000 requests. For comparison, GPT-4o Mini costs 16x less for input tokens.
As of March 2026, the cheapest models for most tasks are Gemini 2.0 Flash ($0.10/1M input), GPT-4.1 Nano ($0.10/1M input), and GPT-4o Mini ($0.15/1M input). The best choice depends on your quality requirements — cheaper models work great for classification, extraction, and simple tasks, while complex reasoning or code generation may need a more capable (and expensive) model.
It depends heavily on the model and volume. A chatbot handling 1,000 conversations per day with GPT-4o Mini costs roughly $15-30/month. The same volume on Claude Opus 4 would cost $1,500+/month. Model selection is the single biggest cost lever — use this calculator to find the sweet spot between quality and budget.
Output tokens are more expensive because the model has to generate them one at a time, running inference at each step. Input tokens are processed in parallel, which is computationally cheaper. That's why output tokens typically cost 2-5x more than input tokens across most providers.
For OpenAI models, we use tiktoken (cl100k_base) — the exact same tokenizer OpenAI uses in production. For other providers, we estimate based on a ratio of approximately 4 characters per token, which is accurate for English text. Actual counts may vary slightly by provider and language.
We update prices regularly based on official provider pricing pages. AI model pricing changes frequently — providers often reduce prices as they optimize infrastructure. Check the provider's official page for real-time rates. Last update: March 2026.
For simple tasks like classification, extraction, or reformatting — use the cheapest models (GPT-4o Mini, Gemini Flash). For complex reasoning, coding, or creative writing — GPT-4o, Claude Sonnet, or Gemini Pro deliver better quality at higher cost. For the most demanding tasks — Claude Opus or o3 provide the highest capability. Use this calculator to compare costs at your expected volume before committing.