AI pricing explained: API token costs, plans and seats
AI pricing is how vendors charge for AI software, and it follows a few models: per-token billing for APIs, monthly subscriptions for apps, per-seat licences for teams, and credits or usage limits on top. This section explains each model, what drives the final bill and how to estimate monthly cost before you commit.
AI costs are harder to predict than most software costs. An API bill depends on how many tokens you send and receive, which model you call and whether you use options such as caching or batch processing. A subscription looks fixed, but plans often cap messages, compute time or features, and team plans add a per-seat charge.
This section explains the mechanics rather than listing prices, because published rates change frequently. You will learn why output tokens usually cost more than input tokens, how context length multiplies spend, when a seat licence beats usage billing, and which hidden costs, such as integration, review time and unused seats, belong in a total-cost estimate.
For current figures, always check the vendor's own pricing page and note the date. The LLM cost calculator in this hub lets you enter the rates you find and estimate a monthly bill, without storing any prices that could go stale.
What you will learn
- How token-based API pricing works, including input, output and cached tokens
- How subscription plans, seat licences and credits compare
- How to estimate a monthly AI bill from expected usage
- Which hidden costs to include in a total-cost-of-ownership estimate
- How to control spend with limits, model choice and prompt design
Beginner lessons
What is a token in AI?
A token is a small piece of text, such as a word, part of a word, a punctuation mark or a space, that a language model reads and writes…
How much does AI cost?
AI can cost nothing or add up to a substantial monthly bill, depending on how it is priced and how much you use it. Consumer assistants…
Key terms
- Token-based pricing
Token-based pricing is a billing model in which an AI provider charges for the number of tokens a model processes, usually quoted per million…
- Seat-based pricing
Seat-based pricing is a billing model in which a software vendor charges a fixed fee per user, or seat, usually per month or year. Many AI business…
- Token
A token is the basic unit of text a language model reads and writes. Tokenizers split text into tokens, which can be whole words, parts of words…
- Context window
A context window is the maximum amount of text, measured in tokens, that a language model can take into account at one time. It includes the system…
- Inference
Inference is the stage where a trained AI model is used to produce output: it takes an input, such as a prompt or an image, and generates a…
- Reasoning model
A reasoning model is a language model trained to work through a problem in intermediate steps before giving its final answer, a process often called…
Explore Mindber
Further reading
- Claude Fable 5: What It Is, How to Use It, and the Prompts That Exploit It
Anthropic's first public Mythos-class model, explained with sources: pricing, safeguards, benchmarks, access methods, real use cases, and copy-paste prompts.
- The True Cost of AI Tools in 2026: Sticker vs Reality
The true cost of AI tools in 2026 runs ~8x the sticker price — a fully-sourced TCO report on LLM API pricing, 7 hidden costs, and how to model it.
- Opus 4.8 Cost Calculator: When It Beats Sonnet & GPT-5.5
Opus 4.8 API cost calculator vs Sonnet 4.6, Haiku 4.5, GPT-5.5. Break-even workloads, smart-routing saves ~32%, per-model cache rates. Any currency.
Frequently asked questions
How is AI API usage priced?
Most language-model APIs charge per token, the small chunks of text a model reads and writes, quoted as a price per million tokens. Input and output tokens are usually priced separately, and output often costs more. Some providers discount cached or batch requests. Your bill is roughly the tokens used multiplied by the rates for the model you call.
Why do output tokens cost more than input tokens?
Generating text takes more computation per token than reading it. A model can process input tokens in parallel, but it produces output one token at a time, with each new token depending on those before it. Providers reflect that difference in their rates, which is why long answers and verbose reasoning can dominate an API bill.
Is a subscription or pay-per-use cheaper?
It depends on volume and pattern. A flat subscription suits a person who uses an assistant heavily every day. Pay-per-use API billing suits applications with variable or low volume, and it scales with actual demand. For teams, compare the per-seat cost with expected usage, and watch for seats that are paid for but rarely used.
How can I estimate my monthly AI cost?
Estimate requests per month, average input and output tokens per request, and the model's rates from the vendor's pricing page. Multiply, then add any fixed subscription or seat fees. Add a margin for retries, longer conversations and growth. The LLM cost calculator in this hub does the arithmetic once you enter current prices.
Last updated