About TokenCost
The honest reference for LLM API pricing.
TokenCost (llmapicosts.com) is a free tool that helps developers and teams understand what large language model APIs actually cost. Per-token pricing is hard to compare across providers — TokenCost normalizes every model to USD per million tokens and turns those rates into a real monthly dollar figure.
What you can do here
- Pricing — browse every current model’s input, output, and cached-input rate on one honest scale, filterable by provider and release year.
- Compare — put models head to head on price, context window, and a shared cost example.
- Calculator — enter your expected traffic and project your monthly bill across every model before you ship.
- Benchmarks — see capability against cost on coding leaderboards (LiveBench, SWE-bench Verified, Aider Polyglot), so you pick the best value, not just the top score.
Where the data comes from
Prices are compiled automatically from live, third-party pricing data (via OpenRouter) and refreshed daily, so the numbers stay current without manual upkeep. Benchmark scores come straight from each benchmark’s own published results (LiveBench and Epoch AI). We explain every source, how prices are normalized, and how the cost-vs-quality join works on our methodology page.
How we keep it accurate
The data pipeline is automated and runs every day, which removes the stale-spreadsheet problem that plagues hand-maintained pricing pages. We normalize every model to a single unit, store each provider’s own cached-input rate rather than guessing, and show the last-updated date openly. Because providers change prices without notice, we treat TokenCost as a planning reference and tell you plainly to confirm the final rate on the provider’s official page before committing spend. When a reader reports an error, we check it against the source and correct it. The full detail on sourcing, verification, and corrections lives on our editorial policy page.
Independence
TokenCost is independent and not affiliated with, sponsored by, or endorsed by any model provider or benchmark. We don’t take payment for placement or ranking. The tools are free to use; the site is supported by advertising.
Why it exists
Choosing an LLM shouldn’t require a spreadsheet and ten open tabs. TokenCost exists to make that decision fast, transparent, and honest — no sign-up, no sales pitch, just the numbers.
Get in touch
Spotted an incorrect price or have a suggestion? We’d love to hear it — head to our Contact page.