We may earn a commission if you buy through links on this page. It never changes a score or a ranking. How we make money
There's a free plan; paid tiers are quoted.
Gemini API offers a Free plan with free input and output tokens, limited model access and generous rate limits. For production, Paid (Pay-as-you-go) has no recurring monthly base fee: you pay for token and request usage at model-specific rates. Grounding and context-cache storage can add separate charges. Google lists some grounding allowances and cache-storage rates, but check the current model pricing and limits before forecasting a workload.
01Every plan, line by line
| Plan | Billed annually | Billed monthly | Built for |
|---|---|---|---|
| Free | — | $0 | Prototyping and limited model access |
| Paid (Pay-as-you-go) | — | Quote | Production apps with metered usage |
Add-ons and overages
- Context caching storage: $0.50 per 1,000,000 tokens per hour through December 31, 2026; $1.00 starting January 1, 2027, plus model-specific cached input token rates.
- Google Search grounding: 5,000 free requests per month shared across Gemini 3 and newer models, then $14 per 1,000 requests; daily cap of 1,500 requests.
- Google Maps grounding: 5,000 free requests per month shared across Gemini 3 and newer models, then $14 per 1,000 search queries.
- Context caching is free on the free plan where supported; Google Search and Maps grounding are unavailable on the free plan.
02What the plan grid doesn't tell you
Model-specific token rates
Paid usage is metered by token or request. Rates vary by model; estimate with the current model pricing before setting a budget.
Grounding after allowance
Search and Maps grounding cost $14 per 1,000 requests or queries after 5,000 free monthly uses shared across Gemini 3 and newer models.
Cache storage
Storage costs $0.50 per million tokens per hour through December 31, 2026, then $1.00 from January 1, 2027, plus cached-input token rates.
Usage caps and pauses
Paid tiers have spend caps, and requests pause when a project reaches its spending limit. Configure limits with production traffic in mind.
03Which plan should you buy?
| If you are… | Get | Why |
|---|---|---|
| Side project prototyping | Free | Free input and output tokens can cover prototyping, but model access and rate limits are limited. |
| Startup in production | Paid (Pay-as-you-go) | There is no monthly base fee, and paid usage has higher production rate limits and spend caps. |
| Multimodal app using grounding | Paid (Pay-as-you-go) | Search and Maps grounding are unavailable on Free; paid usage includes a monthly allowance before per-request fees. |
04Is Gemini API worth the money?
The Free plan is a practical starting point for prototypes that fit its model access and rate limits. For production, Paid (Pay-as-you-go) avoids a recurring base fee, but it does not make usage predictable by itself: token rates depend on the model, and grounding and cache storage can add charges. Set spend caps, monitor requests and budget using the models and features your application will actually call. Self-hosting Gemma is an alternative to evaluate separately, not a way to run Gemini API models locally.
Gemini API pricing FAQ
Does Gemini API have a free tier?
Yes. The Free plan includes free input and output tokens, generous rate limits and limited access to certain models. Context caching is free on the Free plan where supported. Search and Maps grounding are unavailable on the Free plan, and content may be used to improve Google's products.
Does Gemini API charge a monthly subscription?
Paid (Pay-as-you-go) has no recurring monthly base fee. Charges are based on token or request use, with model-specific rates. The plan also includes usage tiers and billing spend caps; requests pause when a project reaches its configured spending limit.
How much does Google Search grounding cost?
For Gemini 3 and newer models, Search grounding includes 5,000 free requests per month, then costs $14 per 1,000 requests. The daily cap is 1,500 requests. Search grounding is unavailable on the Free plan.
What does Gemini API context caching cost?
Context-cache storage costs $0.50 per 1,000,000 tokens per hour through December 31, 2026, and $1.00 starting January 1, 2027. Model-specific cached-input token rates also apply. Context caching is free on the Free plan where supported.