Blog
→
AI Spend Management
October 16, 2026

Tokens to dollars: convert AI tokens to USD

Written by:
Christophe Sion
CEO
Tokens to dollars conversion for AI usage
Share on XShare on LinkedInShare on FacebookShare by email
In this article
Ready to get started?
Get full visibility into your spend in minutes.
Get started for free

Tokens to dollars

AI vendors such as OpenAI, Anthropic and Google bill by the token, a small unit of text, and the price of a single token looks harmless. At $3 per million tokens, one token costs $0.000003. Finance teams usually meet tokens on the invoice, where millions of them add up to a real number that nobody can trace back to a feature or a customer.

Converting tokens to dollars takes one formula: divide the number of tokens by the vendor's pricing unit, usually 1 million, and multiply by the rate. Input tokens (the text you send) and output tokens (the text the model writes back) are priced separately, so you calculate each and add them. This guide walks through the formula with worked examples, explains why your invoice will differ from it, and shows how to turn a token bill into cost per feature and customer.

What is an AI token

An AI token is a small chunk of text, roughly three quarters of a word in English, so 1,000 tokens is about 750 words.

Every request to an AI model is counted in tokens: the question, any instructions or documents sent with it, and the answer that comes back. A token has no fixed dollar value of its own. What it costs depends on the model and on whether it is input or output.

How to convert AI tokens to dollars

Every AI token conversion uses the same formula: cost = token count ÷ pricing unit × rate.

Read the pricing unit first

Most vendors quote prices per 1 million tokens, with separate rates for input and output. Output usually costs more, often several times the input rate. Before running any numbers, check three things on the vendor's pricing page or in your contract:

  • The unit. Per 1 million tokens is standard, but some pages and older contracts quote per 1,000. Mixing the two up makes your estimate wrong by a factor of 1,000.
  • The input and output rates. Calculate each side separately, then add them.
  • The currency. Most AI vendors price in US dollars.

Here is a worked example with illustrative rates of $3 per million input tokens and $15 per million output tokens. A request that sends 1,000 tokens and gets 500 back costs:

  • Input: 1,000 ÷ 1,000,000 × $3 = $0.003
  • Output: 500 ÷ 1,000,000 × $15 = $0.0075
  • Total: $0.0105 per request

An LLM model pricing comparison helps you check that you are comparing rates on the same terms.

Adjust for what the formula misses

The formula gives you the cost at list price. Your invoice will differ because of a few things it leaves out:

  • Cached input. Many vendors charge less for text sent again and again, such as standing instructions, which lowers the input side of the bill.
  • Batch discounts. Requests that can wait for a slower answer are often billed at a lower rate.
  • Reasoning tokens. Some models work through a problem before answering. That hidden thinking is usually billed as output, so the cost can be higher than the visible answer suggests.
  • Token counts per model. Each model splits text into tokens its own way, so the same email can count as more tokens on one model than another. A lower rate per token does not always mean a lower cost per email.
  • Retries. A request that fails and runs again is billed again.
  • Currency. If your company reports in euros, the dollar bill converts at the exchange rate on the payment date, plus any conversion fee your bank or card provider charges.

Move from prompt cost to business cost

A cost of $0.0105 per request looks like nothing. Multiply it by real usage and it becomes a budget line. A feature that handles 20,000 requests a day runs about 600,000 requests a month, which at $0.0105 each comes to $6,300 a month. At an exchange rate of, say, €0.90 to the dollar, that is about €5,670 before conversion fees, for a single feature. Add more features, customers and internal tools, and AI quickly becomes one of the faster-growing costs in the business.

That is why the useful question for finance goes beyond token counts to what AI costs per request, per feature, per customer and per euro of revenue. Founders need that view to protect margin. Finance teams need it to forecast, assign costs to the right owners and stop reconciling several vendor exports at month-end. It is the basis of AI unit economics for SaaS.

Converting tokens to dollars is step one. A practical FinOps for AI framework is step two, where token spend is visible in real time and tied to products, teams, customers and outcomes.

A simple tokens-to-dollars workflow that actually works

A reliable conversion takes four steps:

  1. Get the current rates. Take input and output prices from the vendor's pricing page or your contract, and note the pricing unit.
  2. Pull actual token counts. Use each vendor's usage export or billing data, split into input and output.
  3. Run the formula. Calculate input and output separately, then apply cached-input discounts, batch discounts and currency conversion.
  4. Add business context. Assign each cost to the request, feature, customer or team that created it. That context is also what lets teams set AI budget limits for product teams.

A calculator is enough for a one-off check. For decisions, the conversion needs to run automatically in a dashboard or finance system, because token math done once in a spreadsheet goes stale as soon as usage or prices change. Real-time visibility is what makes AI spend management possible.

How Husk turns token usage into business cost

Husk runs this conversion continuously. It connects to your AI vendors, turns token usage into cost as requests happen, and assigns every euro or dollar to the request, feature, customer and team behind it.

Finance sees AI spend in money, with a customer-level AI P&L that sets each customer's AI cost against their revenue. Budget alerts flag a team or feature heading past its limit, and Husk reconciles each vendor's invoice against actual usage, so the monthly bill matches what the business used. See how Husk's AI spend management works.

Frequently asked questions about token to USD

How much is 1000 tokens in dollars?

At illustrative rates of $3 per million input tokens and $15 per million output tokens, 1,000 tokens cost $0.003 as input or $0.015 as output. That is roughly 750 words of English text. The exact figure depends on the model's rates.

How much is 500 tokens in dollars?

At the same rates, 500 tokens cost $0.0015 as input or $0.0075 as output. If a request has both, calculate each part separately and add them.

How much is 2000 tokens in dollars?

At the same rates, 2,000 tokens cost $0.006 as input or $0.03 as output.

How many tokens equal a dollar?

Divide one dollar by the cost of a single token. At $3 per million, one dollar buys about 333,333 input tokens. At $15 per million, it buys about 66,667 output tokens.

Is one token always worth one dollar?

No. An AI token is a billing unit with no fixed value, and at typical rates a single token costs a tiny fraction of a cent. At $3 per million tokens, one token costs $0.000003.

Why does my token-to-dollar result keep changing?

Usually because the mix shifts: more output relative to input, longer prompts, a different model, or cached and batch discounts applying to some requests and not others. Vendors also change their prices. The formula stays the same while those inputs move.

What is the fastest way to get an accurate conversion?

Take the current rates from the vendor's pricing page or your contract, take actual token counts from your billing data, and apply the formula. If you run a business on AI, go a step further and map that spend to requests, features, customers and revenue as it happens. That is where AI FinOps becomes useful.

‍

Same model, different bill: direct APIs, clouds and routers

The same model can produce a different bill depending on where you buy it, because LLM pricing is several markets stacked on top of each other: model makers selling direct, managed cloud platforms, routing services and hosts that run open models. Each route changes how pricing, access and governance work.

Common buying routes

  • Direct provider API: best for access to the newest models, simpler pricing and fast experimentation. Total cost can rise through separate billing per vendor, fewer controls across providers and no built-in backup if one vendor fails.
  • Managed cloud platform: best for enterprise procurement, control over where data is processed, security and governance. Total cost can rise through platform-specific pricing, regional price differences and possible markup.
  • Router or marketplace: best for switching models quickly, one connection to many vendors and automatic fallback. Total cost can rise through markup, different rounding and extra monitoring or routing fees.
  • Open-model hosting provider: best for low token prices on selected workloads and a wide choice of models. Total cost can rise through uneven quality, missing features and host-specific limits.
  • Self-hosted open models: best for predictable high volume and heavy customization. Total cost can rise through hardware costs, servers sitting idle, engineering time and responsibility for uptime.

Direct provider APIs

Direct APIs are usually the cleanest place to start a comparison. Pricing is easier to read, new models arrive there first, and there is no extra commercial layer between you and the model maker. For testing the newest models or moving quickly, this route gives the clearest view of raw model economics.

Managed cloud platforms

Azure, AWS Bedrock, Vertex AI and similar platforms add considerations beyond price. A company may accept a higher rate because the route fits its procurement process, regional deployment, data-handling policy, network setup or existing cloud commitments, such as an annual spend already agreed with AWS. For many finance and compliance teams that trade-off makes sense, as long as it is measured explicitly.

You deserve
financial clarity.
Get full visibility into your company’s finances in minutes.
Get started for free