Same models. Up to 96% less.

You add prepaid credits to your account. Each request is charged at the model's published per-token rate, in USD per million input and output tokens. Credits are only used when you send requests.

Every model page lists the provider's official API price next to ours. Across 15 models with a published official price, RouterHub is typically 87% cheaper and up to 96% cheaper for the same tokens. For example, 100M input and 10M output tokens on Claude Opus 5.5 cost $600 on Anthropic's API and $108 on RouterHub. Official prices were last checked 2026-11-10.

A typical chat message is about 1,000 input and 500 output tokens. That costs about 0.030¢ on DeepSeek V4.1 Flash, 0.13¢ on Claude Sonnet 5.5 and 0.25¢ on Claude Opus 5.5. Prices for all 26 models are on the models page and in models.json.

No. Creating an account is free, and every message and request is paid from prepaid credits. New accounts get $5 free on their first top-up of $20 or more.

When a provider reports a prompt-cache hit, those input tokens are billed at the lower cached-input rate. Cache writes, where published, are billed at the cache-write rate.

Yes. One prepaid balance pays for web chat and API requests from any tool, at the same per-token rates.

See the Terms for the current credit policy. Your balance and every charge are shown on the Usage and Billing pages.

Top up from Credits & billing in your account. You choose the amount each time.