Google (Gemini) Pricing
Google Gemini models for text, multimodal workloads, and high-throughput inference.
7 models tracked. Rates are per token; output is usually priced higher than input.
| Model | Input / 1M | Output / 1M | Page |
|---|---|---|---|
| Gemini 2.5 Flash | $300.0000(0.0003) | $2500.0000(0.0025) | Open |
| Gemini 2.5 Flash Lite | $100.0000(0.0001) | $400.0000(0.0004) | Open |
| Gemini 3 Flash Preview | $500.0000(0.0005) | $3000.0000(0.003) | Open |
| Gemini 3.1 Flash Image Preview | $300.0000(0.0003) | $2500.0000(0.0025) | Open |
| Gemini 3.1 Flash Lite Preview | $250.0000(0.00025) | $1500.0000(0.0015) | Open |
| Gemini 3.1 Pro Preview | $2000.0000(0.002) | $10000.0000(0.01) | Open |
| Gemini 3.1 Pro Preview Custom Tools | $2000.0000(0.002) | $12000.0000(0.012) | Open |
How to read these rates
Input tokens cover everything you send to the model: system prompt, user message, context, and tool payloads. They are billed at the input rate above.
Output tokens are what the model generates. They are billed at the output rate, which is typically higher because generation is more compute-intensive.
To estimate monthly spend, multiply tokens per request by the relevant rate, then by your request volume. Use the AI Cost Calculator to turn these rates into a budget.