Pricing

Gemini API Pricing in 2026: Every Model, the Free Tier and the Price Jumps Coming

Gemini API pricing per 1M tokens for every model, the free tier, and the promo rates that end Nov 30 and Dec 31, 2026. Here's what your bill does next.

By Engineer Foodie Desk Oct 19, 2026 ✓ Sources checked Oct 19, 2026

6 min read

Gemini API rates from Google’s pricing page (checked Oct 18–19, 2026). USD per 1M tokens, Paid tier, Standard. No affiliate links (how we work).

Google is running promo rates that feel like a gift until the calendar flips. Gemini 4 Argon doubles on Dec 1, 2026. Gemini 3.8 Flash and 3.6 Flash double on Jan 1, 2027. If you’ve been budgeting at the promo line, this is Friends “The One With the Blackout”: the lights go out on cheap tokens, and your bill suddenly needs candles.

For a side-by-side with OpenAI and Claude, compare OpenAI, Claude and Gemini API prices. Deep dives for the other two: OpenAI API pricing and Claude API pricing.

TL;DR by situation

Free / prototype: Use the free tier on Flash, Flash-Lite and 2.5 Pro. Input and output are free of charge; free-tier content is used to improve Google’s products. Gemini 4 Argon is not available on Free.
Cheapest paid: Gemini 2.5 Flash-Lite at $0.10 input / $0.40 output. Gemini 3.1 Flash-Lite is $0.25 / $1.50 if you want a newer Lite.
Top model: Gemini 4 Argon at $2 / $10 through Nov 30, 2026, then $4 / $20 from Dec 1. No free tier. Implicit caching only.
Long prompts: Gemini 2.5 Pro steps up above 200K tokens ($1.25 / $10 → $2.50 / $15). GA Gemini 3.x and 4 list no >200K step.
Jobs that can wait: Batch is 50% off on the paid tier (Flex matches that half-rate on listed models).

Gemini API price table

Paid tier, Standard, USD per 1M tokens. Text rates; audio called out in Notes where Google lists it separately. Verified Oct 18, 2026.

Model	Input	Caching	Output	Notes
Gemini 4 Argon	$2.00 → $4.00	$0.10 → $0.20	$10.00 → $20.00	Doubles Dec 1, 2026. No Free. Implicit cache only
Gemini 3.8 Flash	$0.75 → $1.50	$0.075 → $0.15	$3.75 → $7.50	Doubles Jan 1, 2027. Batch $0.375 / $1.875 (promo)
Gemini 3.6 Flash	$0.75 → $1.50	$0.075 → $0.15	$3.75 → $7.50	Doubles Jan 1, 2027
Gemini 3.5 Flash-Lite	$0.30	$0.03	$2.50	Text / image / video / audio same input
Gemini 3.1 Flash-Lite	$0.25	$0.025	$1.50	Audio input $0.50; audio cache $0.05
Gemini 2.5 Pro	$1.25	$0.125	$10.00	Over 200K: $2.50 / $0.25 / $15.00
Gemini 2.5 Flash	$0.30	$0.03	$2.50	Audio input $1.00; audio cache $0.10
Gemini 2.5 Flash-Lite	$0.10	$0.01	$0.40	Audio input $0.30; audio cache $0.03

Arrow prices are promo → post-promo. Output includes thinking tokens where Google bills them that way.

Two cliffs. Bookmark both.

Gemini 4 Argon after Nov 30, 2026

Through Nov 30, 2026: $2.00 input, $0.10 cached input, $10.00 output. From Dec 1, 2026: $4.00 / $0.20 / $20.00. Exactly 2×.

Worked example: 1M input + 200K output tokens on Argon.

Promo: 1 × $2.00 + 0.2 × $10.00 = $2.00 + $2.00 = $4.00
From Dec 1: 1 × $4.00 + 0.2 × $20.00 = $4.00 + $4.00 = $8.00

Same traffic, double the invoice. Argon is not on the free tier, so there is no free escape hatch.

Gemini 3.8 / 3.6 Flash after Dec 31, 2026

Through Dec 31, 2026: $0.75 input, $0.075 cached, $3.75 output. From Jan 1, 2027: $1.50 / $0.15 / $7.50. Again 2×. Storage for those models also doubles (see caching below).

Worked example: 1M input + 200K output on 3.8 Flash.

Promo: 1 × $0.75 + 0.2 × $3.75 = $0.75 + $0.75 = $1.50
From Jan 1, 2027: 1 × $1.50 + 0.2 × $7.50 = $1.50 + $1.50 = $3.00
Same job on Batch during the promo window: 1 × $0.375 + 0.2 × $1.875 = $0.75

If Flash is your workhorse, January is when the spreadsheet stops lying politely.

Free tier: what’s free and what isn’t

From Google’s pricing page (verified Oct 18, 2026):

Input and output are “Free of charge” for Gemini 3.8 Flash, 3.6 Flash, 3.5 Flash-Lite, 3.1 Flash-Lite, 2.5 Pro, 2.5 Flash and 2.5 Flash-Lite.
Free-tier content is “Used to improve our products”.
Gemini 4 Argon: “not available on the Free tier”.

Fixed free-tier RPM / TPM tables are not published on the rate-limits page (checked Oct 19, 2026). Limits depend on usage tier, “can be viewed in Google AI Studio,” and “Specified rate limits are not guaranteed.” What is printed: RPM, TPM (input) and RPD dimensions; RPD resets midnight Pacific; Tier 1 / 2 / 3 spend windows at $10 / $50 / $200; Batch concurrent requests 100, input file size 2 GB, file storage 20 GB.

So: free tokens yes. Guaranteed free RPM printed on that page: no.

Context caching storage fees

Google does not charge a cache-write fee. You pay the cached-token rate in the table plus storage per 1M tokens per hour:

Models	Storage / 1M tokens / hr
3.8 Flash, 3.6 Flash	$0.50 through Dec 31, 2026; $1.00 from Jan 1, 2027
Flash-Lite line, 2.5 Flash	$1.00
2.5 Pro	$4.50

Gemini 4 Argon supports implicit caching only (no separate write product on the pricing table). Caching pays when you resend a long system prompt or doc pack; storage intensity matters if you hold large caches for hours.

Grounding and Batch

Grounding with Google Search

Verified Oct 19, 2026 on the Gemini pricing page:

Model family	Free tier	Paid
Gemini 3 and newer	Grounding not available on Free for most models	5,000 free search requests/month (shared across Gemini 3+), then $14 per 1,000 requests
Gemini 2.5 Pro	Not available on Free	1,500 RPD free, then $35 / 1,000 grounded prompts
Gemini 2.5 Flash / Flash-Lite	Free: up to 500 RPD (shared)	1,500 RPD free (shared), then $35 / 1,000 grounded prompts

Footnote from Google: one customer request may trigger multiple Search queries; each query is charged. Grounding fees sit on top of token rates.

Batch Mode

On the paid tier, Batch is a 50% cost reduction (confirmed Oct 17 and again Oct 19). Flex matches that half-rate on listed models. Example during the Flash promo window: Gemini 3.8 Flash Batch is $0.375 input / $1.875 output. After the Standard rates double on Jan 1, 2027, Batch stays half of whatever Standard is then.

Gemini API vs Google AI Pro

Different products. The Google AI plans tracker covers consumer Gemini subscriptions: Free, Google AI Plus ($4.99/mo), Google AI Pro ($19.99/mo) and Google AI Ultra (from $99.99/mo), last checked Oct 14, 2026 on Google’s subscription page. Those plans are for chat and consumer features in the Gemini app. API usage is billed separately on the developer pricing page. Paying for Pro does not mean your production traffic is “included.”

If you mostly chat and rarely call the API, a subscription is the calmer bill. If you ship an app, budget tokens (and Grounding, and the Dec/Jan cliffs) on the API meter.

Using the API key with Gemini CLI? Gemini CLI free tier limits covers the CLI’s free-tier changes.

Getting 429s through Gemini CLI on an API key? Gemini CLI “429 RESOURCE_EXHAUSTED” explains the quota behind it.

FAQ

Is the Gemini API free? Yes for listed Flash, Flash-Lite and 2.5 models on Free; Argon is not. Free-tier content is used to improve Google’s products.

When do Gemini prices go up? Argon doubles Dec 1, 2026; 3.8 and 3.6 Flash (and their storage rates) double Jan 1, 2027.

What’s the cheapest Gemini model? Paid: 2.5 Flash-Lite at $0.10 / $0.40. Free tier: $0 on the models Google marks free of charge.

Does Google AI Pro include API access? No. Consumer Google AI plans and the Gemini API are billed separately.

Is Batch half price? Yes on the paid tier: 50% off Standard. Flex matches Batch on listed models.

Are free-tier RPM limits published? Not as a fixed RPM/TPM table on the rate-limits page; check Google AI Studio for your tier.

Sources

Gemini API pricing: https://ai.google.dev/gemini-api/docs/pricing (model table, free-tier wording, storage, promo cliffs checked Oct 18, 2026; Grounding and Batch confirmed Oct 19, 2026)
Gemini API rate limits: https://ai.google.dev/gemini-api/docs/rate-limits (checked Oct 19, 2026; fixed free RPM/TPM table not published)
Google AI consumer subscriptions: https://gemini.google/us/subscriptions/?hl=en (checked Oct 14, 2026 via our tracker)