AI API Pricing Comparison and Token Costs

Compare input, output, cache, per-call, and official reference rates for models available on API-Route, then estimate AI API token costs.

API-Route model rates, cache rates, and official price references

Text prices are shown per 1M tokens. Media models use the call, second, or specification unit shown in the table.

API-Route model rates, cache rates, and official price references
ModelInputOutputCache readCache creationBilling unit
gpt-5.6-sol$0.253676$1.5221$0.025368$0.317096USD per 1M tokens
claude-sonnet-4-6$0.242647$1.2132$0.024265$0.303309USD per 1M tokens
gemini-3.1-pro$0.404412$2.4265$0.404412$0.505515USD per 1M tokens
deepseek-v4-pro$0.882353$1.7647$0.007353$0USD per 1M tokens
grok-4.5$0.091912$0.275735$0.022978$0.11489USD per 1M tokens
qwen/qwen3.5-plus-20260420$0.525$3.15$0.525$0.65625USD per 1M tokens

Related pages

AI API Plans and Packages

Packages are optional: top up your account and use the API with pay-as-you-go billing, or compare daily, weekly, monthly, and quota-based plans when bundled quota fits your usage.

API-Route FAQ

Answers about OpenAI-compatible Base URL, API keys, model names, LibreChat, Codex, Claude Code, VS Code, plans, payments, and white-label AI API reseller platform setup.

FAQ

How do I compare OpenAI, Claude, and Gemini API pricing?

Start with input and output token rates, then include cache pricing, per-call models, context length, and expected request volume.

How do I estimate AI API token cost?

For text models, estimate input token cost plus output token cost plus cache-related costs. Image, audio, and video models follow the displayed specification or per-call price.

Why include official reference pricing?

Official references help compare public provider prices. Actual billing follows API-Route rates, account records, and usage logs.