Direct answer: Gemini 3 Pro API has no current rate
Gemini 3 Pro Preview is no longer available through the Gemini API: Google says it was shut down on March 9, 2026. So there is no current per-token price to use for new API calls to that model. Google’s shutdown notice recommends migrating to Gemini 4 Argon. (Google model documentation)
For historical context only, the third-party pricing comparison by Solvimon lists Gemini 3 Pro Preview at $2 per 1 million input tokens and $12 per 1 million output tokens, in a timeline entry dated November 18, 2025. Treat those as reported historical rates—not current Google pricing or a production quote. (Solvimon)
Pricing status checked November 10, 2026.
Don’t confuse Gemini 3 Pro with Gemini 3.1 Pro
The models have similar names, but the available rate information refers to different models. Solvimon’s July 13, 2026 pricing snapshot lists Gemini 3.1 Pro Preview at $2 input and $12 output per million tokens for prompts up to 200,000 tokens, and $4/$18 above that threshold. Those figures—and the stated threshold—are for Gemini 3.1 Pro Preview, not Gemini 3 Pro. (Solvimon)
For a budget, use Google’s pricing page and documentation for the exact model and API route you plan to call. Google’s pricing documentation separates free access from paid production usage; it lists higher rate limits and features such as caching and Batch API for paid use. The applicable price and terms depend on the model and route. (Google AI Studio pricing)
How to estimate API spend
For token-priced text usage, calculate input and output separately:
Estimated token cost = (input tokens ÷ 1,000,000 × input rate) + (output tokens ÷ 1,000,000 × output rate)
For example, using the historically reported Gemini 3 Pro rates above, a hypothetical month with 30 million input tokens and 30 million output tokens works out to:
30 × $2 + 30 × $12 = $420
That arithmetic illustrates how input and output rates affect a bill. It is not a current Gemini 3 Pro estimate: Google shut down Gemini 3 Pro Preview on March 9, 2026. Separately, Solvimon’s $420 monthly workload example is for Gemini 3.1 Pro, not Gemini 3 Pro. (Google model documentation; Solvimon)
For a replacement model, measure representative traffic from your own application, project monthly input and output tokens, and apply the current rate for that exact model. Add cache, batch, search, or multimodal charges only if the current pricing terms confirm they apply to your model and route. Google’s billing documentation also describes account tiers and billing setup; API costs are usage-based, not a consumer chat subscription fee. (Google billing documentation)
Production budget checklist
Before committing to a forecast, confirm:
- Exact model ID and availability: Gemini 3 Pro Preview is shut down; choose a currently available model instead. (Google model documentation)
- Official rate card: Record the model-specific input and output prices, currency, billing unit, and any context-length tiers from Google’s pricing page. (Google AI Studio pricing)
- Route and provider: Confirm whether you are calling Google directly or using an intermediary, since another provider’s endpoint and terms are not automatically Google’s.
- Realistic token volumes: Estimate from representative request logs, including the prompt and response sizes your application actually produces.
- Other billable usage: Check whether caching, batch processing, search grounding, or image/audio/video usage changes the total for your chosen model.
Bottom line: Solvimon reports Gemini 3 Pro Preview’s historical rate as $2/$12 per million input/output tokens, but it is not a usable current price because Google shut the model down. For production budgeting, select an available model and use its current, model-specific Google rate card.
Sources
Gemini 3 Pro preview — Gemini API | Google AI for Developers
Google Gemini pricing vs OpenAI: API cost comparison