Some Gemini models have free-tier access, but not all of them. Google's pricing and rate-limit docs say the exact limits depend on the specific model and your usage tier, and should be checked in AI Studio.

What is the cheapest Gemini model right now?

In Google's current public pricing page, Gemini 3.1 Flash-Lite Preview is the cheapest text model. Standard pricing is $0.25 per million input tokens and $1.50 per million output tokens. Batch and Flex pricing are $0.125/$0.75.

How much does Gemini 2.5 Pro cost?

Gemini 2.5 Pro uses prompt-size-based pricing. Standard pricing is $1.25 input and $10 output per million tokens for prompts up to 200K tokens, and $2.50 input and $15 output above 200K tokens.

Does Gemini rate limiting use fixed RPM numbers?

No. Google's current rate-limit docs say limits vary by model and usage tier, preview models are more restricted, and your active limits should be checked in AI Studio.

How do Batch and Flex compare to Standard pricing?

For the models covered here, Batch and Flex are generally about 50% cheaper than Standard pricing. If your workload can tolerate asynchronous or lower-priority processing, those cheaper tiers matter more than micro-optimizing prompts.

Gemini API Pricing: Current Flash, Flash-Lite, and Pro Rates (April 2026)

Google's Gemini pricing is more nuanced than a single provider table can capture. The current official docs split pricing by model, service tier (Standard, Batch, Flex, Priority), and in some cases prompt size. That is why older one-line summaries often get Gemini wrong.

This guide uses the current official Gemini pricing page and Gemini rate-limit page.

The key pricing rows to know

Gemini 3.1 Pro Preview

Tier	Input $/M	Output $/M	Notes
Standard	$2.00 <= 200K prompt, $4.00 > 200K	$12.00 <= 200K, $18.00 > 200K	No free tier listed
Batch	$1.00 <= 200K, $2.00 > 200K	$6.00 <= 200K, $9.00 > 200K	Half-price async tier
Flex	$1.00 <= 200K, $2.00 > 200K	$6.00 <= 200K, $9.00 > 200K	Lower-priority tier

Gemini 3.1 Flash-Lite Preview

Tier	Input $/M	Output $/M
Standard	$0.25	$1.50
Batch	$0.125	$0.75
Flex	$0.125	$0.75

Google explicitly describes Flash-Lite as its most cost-efficient model and notes that preview models may change before becoming stable.

Gemini 3 Flash Preview

Tier	Input $/M	Output $/M
Standard	$0.50	$3.00
Batch	$0.25	$1.50
Flex	$0.25	$1.50

Gemini 2.5 Flash

Tier	Input $/M	Output $/M
Standard	$0.30	$2.50
Batch	$0.15	$1.25
Flex	$0.15	$1.25

Gemini 2.5 Pro

Tier	Input $/M	Output $/M	Notes
Standard	$1.25 <= 200K prompt, $2.50 > 200K	$10.00 <= 200K, $15.00 > 200K	Free tier listed
Batch	$0.625 <= 200K, $1.25 > 200K	$5.00 <= 200K, $7.50 > 200K	Half-price async tier
Flex	$0.625 <= 200K, $1.25 > 200K	$5.00 <= 200K, $7.50 > 200K	Lower-priority tier

What changed versus the stale summaries still circulating

Three things matter:

Gemini 3.1 Pro Preview is not a $1.25/$5 model. The current official page lists it materially higher.
Gemini 2.5 Pro pricing changes above 200K prompt tokens. If you ignore that threshold, you under-estimate cost.
Google does not publish one universal free-tier RPM number. The rate-limit docs say limits vary by model and usage tier, preview models are more restricted, and your live limits should be checked in AI Studio.

Real cost example: one workload across cheap, mid, and premium Gemini tiers

Assume 5,000 requests per day, each with 2,500 input tokens and 400 output tokens.

Gemini 3.1 Flash-Lite Preview Standard

Daily input: 5,000 x 2,500 x $0.25 / 1M = $3.125
Daily output: 5,000 x 400 x $1.50 / 1M = $3.00
Monthly total: about $183.75

Gemini 2.5 Flash Standard

Daily input: 5,000 x 2,500 x $0.30 / 1M = $3.75
Daily output: 5,000 x 400 x $2.50 / 1M = $5.00
Monthly total: about $262.50

Gemini 2.5 Pro Standard

This prompt is well under 200K tokens, so the lower Pro bracket applies:

Daily input: 5,000 x 2,500 x $1.25 / 1M = $15.625
Daily output: 5,000 x 400 x $10.00 / 1M = $20.00
Monthly total: about $1,068.75

Gemini 3.1 Pro Preview Standard

Daily input: 5,000 x 2,500 x $2.00 / 1M = $25.00
Daily output: 5,000 x 400 x $12.00 / 1M = $24.00
Monthly total: about $1,470

That is the real current shape of Gemini pricing: cheap Flash tiers, a middle 2.5 Pro tier, and a more expensive 3.1 Pro Preview tier.

When to use which Gemini tier

Use Gemini 3.1 Flash-Lite Preview when minimizing cost matters most.
Use Gemini 2.5 Flash when you want a cheap general-purpose workhorse without preview-only positioning.
Use Gemini 2.5 Pro when you need Google's stronger reasoning and coding tier without jumping to 3.1 Pro Preview pricing.
Use Gemini 3.1 Pro Preview when you specifically want the newest Pro-preview model and are willing to pay for it.

Rate-limit advice that is actually correct

Google's current rate-limit docs say:

limits are measured across RPM, TPM, and RPD
limits vary by model and usage tier
preview models are more restricted
your active limits should be checked in AI Studio

So do not hard-code old blog-post numbers like "Flash is 15 RPM and Pro is 2 RPM" into planning docs. That is not how Google documents Gemini limits today.

The practical takeaway

For Gemini, cost control is mostly about tier selection and service tier selection:

Pick the right model.
Use Batch or Flex when latency is not critical.
Watch the 200K prompt threshold on Pro models.
Verify current quotas in AI Studio, not in screenshots or stale blog posts.

For a multi-provider comparison, see our LLM pricing overview.