GoogleUpdated 2h ago

Google: Gemini 3.1 Pro Preview Custom Tools API Pricing

Live token cost for Google: Gemini 3.1 Pro Preview Custom Tools from Google. Use the figures below for budgeting, then tune your exact request mix in the interactive calculator. Prices refresh every 24 hours from OpenRouter.

Input
$2.00
/ 1M tokens
Output
$12.00
/ 1M tokens
Cached input
$0.200
/ 1M tokens

Capabilities

1049K context66K max outputVisionPrompt cachingReasoning tokens

Google: Gemini 3.1 Pro Preview Custom Tools cost at scale

Estimated monthly cost across common production volumes. Assumes 30-day months and the request shapes shown.

TierRequests / dayIn / out tokens$ / month
Hobby1,000500 / 200$102.00
Startup10,0001,500 / 500$2,700.00
Growth100,0003,000 / 800$46,800.00
Enterprise1,000,0008,000 / 2,000$1,200,000.00
Open Google: Gemini 3.1 Pro Preview Custom Tools in interactive calculator →

Compare Google: Gemini 3.1 Pro Preview Custom Tools vs.

Frequently asked questions

How much does Google: Gemini 3.1 Pro Preview Custom Tools cost?

Google: Gemini 3.1 Pro Preview Custom Tools costs $2.00 per 1M input tokens and $12.00 per 1M output tokens, with cached input at $0.20 per 1M tokens. A typical 1,500-token in / 500-token out request costs $0.00900.

Does Google: Gemini 3.1 Pro Preview Custom Tools support cached input?

Yes. Google: Gemini 3.1 Pro Preview Custom Tools supports prompt caching at $0.20 per 1M cached input tokens. Reuse the same system prompt or context across requests to cut input cost dramatically.

What is the Google: Gemini 3.1 Pro Preview Custom Tools context window?

Google: Gemini 3.1 Pro Preview Custom Tools supports a context window of 1,048,576 tokens (1049K). Max output per response is 65,536 tokens.

What is Google: Gemini 3.1 Pro Preview Custom Tools good for?

Google: Gemini 3.1 Pro Preview Custom Tools is a good fit for long context (1M+ tokens), multimodal grounding, document QA. For other use cases, run your specific input/output mix through the interactive calculator to compare against alternative models.