Short answer: Third-party pricing references report Gemini 2.5 Pro at $1.25 per million input tokens and $10 per million output tokens for prompts up to 200,000 tokens. They report $2.50 per million input tokens and $15 per million output tokens above that threshold. Treat these as reported reference rates, not a confirmed live Google quote; check Google’s Gemini Developer API pricing page before budgeting production use. Convly and Chapter.pub report the same figures.
Gemini 2.5 Pro API rates
| Reported prompt tier | Input price | Output price |
|---|---|---|
| Up to 200,000 tokens | $1.25 / 1M tokens | $10.00 / 1M tokens |
| Above 200,000 tokens | $2.50 / 1M tokens | $15.00 / 1M tokens |
The third-party references describe the threshold in terms of prompt size or length. One says the higher rates apply to the entire request when the prompt exceeds 200,000 tokens, rather than only to tokens above the threshold. That is a reported rule, not a billing condition confirmed here by Google. Verify how the threshold is measured and applied on the pricing page before relying on a long-prompt estimate. RapidDevelopers
How to estimate API token costs
For a token-only estimate, calculate input and output separately:
Estimated token cost = (billable input tokens ÷ 1,000,000 × input rate) + (billable output tokens ÷ 1,000,000 × output rate)
For example, assuming a prompt below the reported threshold and 100,000 input tokens plus 20,000 output tokens:
- Input: 100,000 ÷ 1,000,000 × $1.25 = $0.125
- Output: 20,000 ÷ 1,000,000 × $10 = $0.20
- Illustrative token-only total: $0.325
Google’s thinking guide says pricing is based on the full thought tokens, so visible answer length alone may understate the output-related cost of a reasoning request. Use the returned usage information to estimate generated tokens, including thinking tokens; confirm the applicable usage field and rate on Google’s current pricing and API documentation. Google also describes Gemini 2.5 Pro as a thinking model for complex code, math, STEM, and long-context analysis. Model documentation
For a prompt above 200,000 tokens, the same third-party references report higher rates. If the reported whole-request rule applies, an example with 250,000 input tokens and 20,000 output tokens would be:
- Input: 250,000 ÷ 1,000,000 × $2.50 = $0.625
- Output: 20,000 ÷ 1,000,000 × $15 = $0.30
- Conditional illustrative token-only total: $0.925
This example depends on reported rates and an unconfirmed threshold rule; it is not a Google quote or a monthly forecast. Your actual budget depends on usage and the rates applicable to your route. Add separately priced features only after confirming their current price and eligibility.
API access and the December 1 routing notice
Google’s Gemini 2.5 Pro API model page says access to Gemini 2.5 models is limited to users who have actively used them in the past. It also says the models will continue to be served through the API until further notice and recommends newer models for new projects. If you have not actively used Gemini 2.5 Pro before, do not assume you can start a new integration at the reported rates.
Separately, an AI Studio deprecation notice states that, effective December 1, 2026, API requests to gemini-2.5-pro will automatically route to gemini-3.8-flash and be billed at Gemini 3.8 Flash rates. That date is in the future as of November 10, 2026. The model page’s “until further notice” statement and the dated routing notice do not, by themselves, resolve how the change applies to every existing integration. Check the current notices and your project’s status before planning around the model ID or a particular rate.
This page concerns the Gemini Developer API. It does not establish prices or access terms for Google Cloud’s separate platform offerings or third-party gateways. The reported token prices are not consumer Gemini subscription prices, and subscription or free access should not be assumed to provide the same production API availability or billing.
Bottom line: Secondary references report rates of $1.25/$10 per million input/output tokens up to a 200,000-token prompt and $2.50/$15 above it. Google’s documentation says thinking tokens are included in pricing, but confirm the current applicable rates, threshold billing, access eligibility, and route before committing a production budget.
Sources
Gemini Developer API pricing — Google AI Studio
Gemini 2.5 Pro — Gemini API | Google AI for Developers
Gemini thinking — generateContent API
Gemini deprecations — Gemini API | Google AI Studio
Gemini 2.5 Pro Pricing: API Cost per 1M Tokens