Plan & Cost

AI Cost Calculator

Compare the monthly cost of up to four models on the same workload, using the verified prices from the Model Explorer.

Start from:

Currency

Workload

Requests per month

Input tokens / requestPrompt, context and retrieved textOutput tokens / request

Cached input%Share of input tokens served from the provider's prompt cache

Models to compare

2 of 4

Model 1

Claude Sonnet 5.5 — AnthropicFirst-party

Model 2

Gemini 2.5 Flash — GoogleFirst-party

Lowest monthly cost

$17

Gemini 2.5 Flash

Annual at lowest

$204

10,000 requests / month

Monthly cost by model

Model	Per request	Monthly	vs lowest
Claude Sonnet 5.5Anthropic · $2.00 / 1M tokens in · $10.00 / 1M tokens out	$0.008	$80	+$63
Gemini 2.5 FlashGoogle · $0.30 / 1M tokens in · $2.50 / 1M tokens out	$0.0017	$17	Lowest

Would self-hosting be cheaper at this volume?API vs Local Turn this into a full business case for a clientBusiness Case Builder

Lowest monthly cost · Gemini 2.5 Flash

$17

See results

Assumptions & sources

Formula per request: input tokens × (1 − cached %) × input rate + input tokens × cached % × cached-input rate + output tokens × output rate. Monthly = per request × requests; annual = monthly × 12.

Prices are each model's current verified list price from the catalog, linked from the table. Models without a verified cached-input price charge cached tokens at the standard input rate. Presets are illustrative starting points, not benchmarks.

Not included: retries, tool-call overhead, embeddings, reranking, hosting and engineering time. Results are estimates, not a provider invoice.