Simple, pay-as-you-go pricing

You pay the official list price of each model, per token. No subscription, no minimum spend, no hidden fees.

Official list prices

Per-model prices match the providers' published list prices, in USD.

Pay only for usage

Charged per request based on input, output and cached tokens actually used.

One balance for all models

Top up once and use it across every model in the catalog.

Transparent billing

Every request is itemized in the console with tokens and cost.

How a request is billed

cost = input tokens × input price + output tokens × output price + cached tokens × cache price

Prices are per 1M tokens. Cached prompt tokens are billed at the (much lower) cache read price. Some models use tiered pricing by input length; image and media models may be billed per request.

List prices

Official list prices in USD. Token prices are per 1M tokens.
ModelTypeInputOutputCache readContext
gpt-5.6-terraOpenAIChat$2$12$0.2—
gpt-5.6-lunaOpenAITieredChat$0.2$1.2$0.02—
gpt-5.6-solOpenAIChat$5$30$0.5—
gpt-5.5OpenAIChat$5$30$0.5—
gpt-5.2OpenAIChat$1.75$14$0.175—
gpt-5.3-codexOpenAIChat$1.75$14$0.175—
gpt-4.1OpenAIChat$2$8$0.5—
gpt-5.4OpenAIChat$2.5$15$0.25—
claude-sonnet-4-6AnthropicChat$3$15$0.3—
claude-opus-4-8AnthropicChat$5$25$0.5—
gpt-5.4-miniOpenAIChat$0.75$4.5$0.075—
deepseek-v4-flashDeepSeekChat$0.14$0.28$0.03—
gpt-5.4-nanoOpenAIChat$0.2$1.25$0.02—
gpt-5.4-proOpenAIChat$30$180$30—
claude-opus-4-6AnthropicChat$5$25$0.5—
claude-opus-4-7AnthropicChat$5$25$0.5—

Billing FAQ

Do you charge anything on top of the list price?+

No. The prices shown here are what you are billed per token. There are no platform or subscription fees.

How do I pay?+

Top up your balance in the console. Requests are deducted from the balance in real time.

Are failed requests billed?+

No. Requests that fail with an error are not charged.

Do you offer volume pricing?+

For large or committed volumes, contact us through the console and we will work out a plan.