AI Price Metrics

Cost calculator

Estimate what a given token volume will cost on a specific model.

GPT-6.1 Sol (OpenAI) — estimated cost

For 1,000,000 input + 1,000,000 output tokens

Input cost$2.00
Output cost$10.00
Total$12.00

How AI token pricing works

Most AI text models are billed by the token, and providers quote prices per million tokens. Input tokens are everything you send — your prompt, instructions, documents and conversation history — and output tokens are what the model writes back. The two are priced separately, and output usually costs more.

To estimate a bill, divide each token count by one million and multiply by that side of the price: input tokens × input price, plus output tokens × output price. On a model that charges $2.00 per million input tokens and $10.00 per million output tokens, one million of each costs $12.00. Image, video and audio models are priced per image, second, minute or character instead, so they are listed on the comparison table rather than here.

Cheapest text models per million tokens

Ranked on the cost of 1M input plus 1M output tokens. Open a model for its full price history.

ModelInput / 1MOutput / 1M
Command R7B · Cohere$0.0375$0.15
Gemini 2.0 Flash-Lite · Google$0.075$0.30
Gemini 2.0 Flash · Google$0.10$0.40
GPT-6 Luna · OpenAI$0.10$0.50
GLM 5.3 Flash · Perplexity$0.15$0.50
Command R · Cohere$0.15$0.60
GPT-5.6 Luna · OpenAI$0.20$1.20
DeepSeek V4.1 Flash · DeepSeek$0.30$1.20
Gemini 3.1 Flash-Lite · Google$0.25$1.50
DeepSeek V4 Flash · DeepSeek$0.44$1.32

Frequently asked questions

What is a token?

A token is a small chunk of text that a model reads or writes — in English, roughly four characters or three quarters of a word, so 1,000 tokens is about 750 words.

Why do input and output tokens cost different amounts?

Writing text takes more computation than reading it, so providers usually charge several times more per output token than per input token. The calculator prices each side separately.

Does the estimate include caching, batch discounts or tool fees?

No. It uses each model's standard per-token rates as tracked on this site. Discounts for cached input or batch jobs, and extra charges for tools such as web search, are not included — check the provider's pricing page for those.

How current are the prices?

As current as our data: see the About page for how prices are collected and refreshed. Results are in US dollars; the currency selector converts them for reference only.

We use cookies for analytics, and — once ads are enabled — to personalize them. See our Privacy Policy.