AIPrice / Global edition
Know the cost before you choose the model.
Compare OpenRouter's listed model rates, then put your own token volume through the calculator.
OpenRouter catalog snapshot: · Listed rates, not a live quote
Monthly cost estimate
Use the same workload across every model. Listed text-token rates are shown in USD.
Loading reference prices…
Text input and output only. Published prompt-length tiers are applied when listed. Caching, tools, media, taxes and routing may change the bill.
02 / BENCHMARK + PRICE
A value shortlist, with a visible rule.
Artificial Analysis Intelligence Index ÷ estimated monthly cost. The cards recalculate with your workload above; the default is 10,000 requests with 1,500 input and 500 output tokens each. Only paid models with a published score of at least 30 appear.
Loading scored models…
Low listed token rates and a 1M-token context window. OpenRouter does not publish a comparable Intelligence Index for this model in this snapshot, so it is left out of the value ranking.
This ratio is a screening aid, not a test of quality on your task. Artificial Analysis' Intelligence Index primarily tests English-language text tasks. “US developers” describes selected model authors, not where your data is processed. A listed model may be unavailable to your account or location under provider terms; check access before relying on a ranking.
Explore OpenRouter's Value leaders03 / THE MODEL CATALOG
Find a fit for your budget.
Explore — selected text-output models from OpenRouter's weekly-use ordering. Select a row to estimate its cost.
| Model / developer | AA Intel. | Input | Output | Context | Your month | Source | Use in calculator |
|---|
No models match those filters. Try a broader search.
Rates are OpenRouter catalog list prices in the stored snapshot, not live quotes. Free routes can have usage limits. Follow each model link for current endpoints, availability and price tiers.
04 / READ THE DATA
A useful estimate starts with clear assumptions.
Official catalog snapshot
Rates and available benchmark fields came from OpenRouter's model API on . The list selects popular text-output models; usage is not a quality measure.
One workload
Monthly estimate = requests × (input tokens × input rate + output tokens × output rate) / 1,000,000. Published prompt-length rate tiers are applied per request. Other charges are excluded.
Score and price
The shortlist uses the Artificial Analysis Intelligence Index shown in OpenRouter metadata, with a score floor of 30. Its method primarily tests English-language text tasks. Missing scores stay unranked. Test candidates on your task before spending.