Gemini API Pricing 2026: Every Model, the Free Tier and the Cheapest Provider
Official Gemini API prices for Gemini 4, 3.8 Flash, 3.1 Pro, 2.5 and Gemma, what the free tier covers, Batch, Flex and caching discounts, and which providers sell Gemini for less.
Gemini API prices run from $0.10 input / $0.40 output per million tokens (Gemini 2.5 Flash-Lite) to $2 / $10 for Gemini 4 Carbon and Argon, which rise to $4 / $20 on December 1, 2026. Most Flash and Flash-Lite models, Gemini 4 Hydrogen and Gemma 4 can be used free in Google AI Studio; Gemini 4 Carbon, Argon and 3.1 Pro can’t. Past the free tier, the cheapest options are Google’s own Batch and Flex tiers (50% off) or a reseller such as Kie.ai, which sells Gemini 3.8 Flash at 30% of Google’s price.
Official figures below are from Google’s Gemini API pricing page, checked on November 17, 2026, and tracked on our Google provider page. Third-party prices come from our provider listings on the same date.
Gemini API pricing at a glance
Standard (real-time) prices per 1M tokens on the Gemini Developer API. Output prices include thinking tokens.
| Model | Input | Output | Free tier |
|---|---|---|---|
| Gemini 4 Carbon | $2.00 ($4.00 from Dec 1) | $10.00 ($20.00 from Dec 1) | No |
| Gemini 4 Argon | $2.00 ($4.00 from Dec 1) | $10.00 ($20.00 from Dec 1) | No |
| Gemini 4 Hydrogen | $1.50 | $7.50 | Yes |
| Gemini 3.8 Flash | $0.75 ($1.50 from Jan 1) | $3.75 ($7.50 from Jan 1) | Yes |
| Gemini 3.6 Flash | $0.75 ($1.50 from Jan 1) | $3.75 ($7.50 from Jan 1) | Yes |
| Gemini 3.5 Flash-Lite | $0.30 | $2.50 | Yes |
| Gemini 3.1 Pro Preview | $2.00 / $4.00 | $12.00 / $18.00 | No |
| Gemini 3.1 Flash-Lite | $0.25 ($0.50 audio) | $1.50 | Yes |
| Gemini 3 Flash Preview | $0.50 ($1.00 audio) | $3.00 | Yes |
| Gemini 2.5 Pro | $1.25 / $2.50 | $10.00 / $15.00 | Yes |
| Gemini 2.5 Flash | $0.30 ($1.00 audio) | $2.50 | Yes |
| Gemini 2.5 Flash-Lite | $0.10 ($0.30 audio) | $0.40 | Yes |
| Gemini Embedding 2 | $0.20 (text) | – | Yes |
Where a row shows two prices separated by a slash, the first applies to prompts up to 200K tokens and the second to longer prompts. Only Gemini 3.1 Pro and 2.5 Pro price by prompt length; the other models, including the 1M-context Gemini 4 models, charge one rate for any prompt length.
Two scheduled price rises matter if you are budgeting now. Gemini 4 Carbon and Argon double on December 1, 2026, and Gemini 3.8 Flash and 3.6 Flash double on January 1, 2027. Gemini 4 Carbon, Argon and Hydrogen are not in our listings yet, so they have no model pages.
How much does Gemini 4 cost?
Gemini 4 comes in three models. Carbon is Google’s most capable reasoning model, and Argon is the faster Gemini 4 reasoning model that uses fewer tokens. Both cost $2 / $10 per million tokens until November 30, then $4 / $20. Hydrogen, aimed at everyday coding and agent work, costs $1.50 / $7.50 with no announced increase, and it is the only Gemini 4 model on the free tier.
Gemini 4 supports only implicit caching: repeated prompt prefixes are billed at the cached rate automatically ($0.10 per 1M tokens for Carbon and Argon until December 1, $0.075 for Hydrogen), and you can’t create explicit caches.
Is the Gemini API free?
Yes, for most models. The free tier gives free input and output tokens on Gemini 4 Hydrogen, 3.8 Flash, 3.6 Flash, 3.5 Flash-Lite, 3.1 Flash-Lite, 3 Flash Preview, 2.5 Pro, 2.5 Flash, 2.5 Flash-Lite, Gemini Embedding 2 and Gemma 4. Gemini 4 Carbon, Gemini 4 Argon, Gemini 3.1 Pro and the image-generation models are paid only.
The limits that apply:
- Rate limits are per project, not per API key, so creating extra keys doesn’t give you more quota.
- Daily request quotas reset at midnight Pacific time.
- Google no longer publishes a free-tier limits table. You see your project’s actual RPM, TPM and RPD limits in AI Studio.
- Free-tier prompts and responses may be used to improve Google’s products. Paid-tier data is not.
- No Batch API on the free tier. Batch, higher rate limits and the paid models need a billing account.
Paid usage starts at Tier 1 as soon as you link a billing account. Tier 2 requires $100 of spend and 3 days since your first payment, and Tier 3 requires $1,000 and 30 days. Each tier raises your rate limits and spending caps.
For free tiers across other providers, see our AI API free tier guide.
Gemini Batch, Flex and context caching discounts
- Batch API: 50% off input and output for asynchronous jobs. Gemini 3.8 Flash drops to $0.375 / $1.875 and Gemini 3.1 Pro to $1 / $6.
- Flex: same price as Batch, for real-time requests that accept slower, lower-priority processing.
- Priority: 1.8x the standard price for higher reliability (Gemini 3.8 Flash costs $1.35 / $6.75).
- Context caching: cached tokens cost 10% of the input price (Gemini 3.8 Flash $0.075 per 1M tokens), plus an hourly storage fee on explicit caches ($0.50 to $4.50 per 1M tokens per hour depending on the model).
- Google Search grounding: on Gemini 3 and newer, 5,000 grounded requests a month are free across all models, then $14 per 1,000.
Gemma API pricing
Gemma is Google’s open-weight model family. On the Gemini API, Gemma 4 is free tier only: input, output and caching are free, and there is no paid tier, so production traffic usually goes to a third-party host. Cheapest listings we track:
| Model | Cheapest | Next cheapest |
|---|---|---|
| Gemma 4 31B | $0.09 / $0.34 (Infron, fp4) | $0.10 / $0.34 (OpenRouter, fp4) |
| Gemma 4 26B A4B | $0.017 / $0.052 (OpenRouter) | $0.07 / $0.34 (DeepInfra) |
| Gemma 3 27B | $0.064 / $0.128 (OpenRouter, fp8) | $0.08 / $0.16 (DeepInfra) |
| Gemma 3 12B | $0.04 / $0.12 (OpenRouter) | $0.05 / $0.10 (Novita AI) |
| Gemma 3 4B | $0.04 / $0.08 (OpenRouter) | $0.05 / $0.10 (DeepInfra) |
The cheapest Gemma endpoints often run quantized weights (fp4 or fp8), so check the precision if output quality matters. Cloudflare Workers AI hosts Gemma 4 26B A4B at $0.10 / $0.30 and includes 10,000 free Neurons a day. Gemma 4 12B has a model page but no paid listings yet.
Which provider sells Gemini cheapest?
Real-time prices per 1M tokens, input / output:
| Model | Cheapest reseller | Cheapest Flex route | |
|---|---|---|---|
| Gemini 3.8 Flash | $0.75 / $3.75 | $0.225 / $1.125 (Kie.ai) | $0.30 / $1.50 (Infron) |
| Gemini 3.6 Flash | $0.75 / $3.75 | $0.225 / $1.125 (Kie.ai) | $0.30 / $1.50 (Infron) |
| Gemini 3.5 Flash-Lite | $0.30 / $2.50 | $0.24 / $2.00 (Infron) | $0.12 / $1.00 (Infron) |
| Gemini 3.1 Pro | $2.00 / $12.00 | $1.60 / $9.60 (Infron) | $0.80 / $4.80 (Infron) |
| Gemini 3.1 Flash-Lite | $0.25 / $1.50 | $0.20 / $1.20 (Infron) | $0.10 / $0.60 (Infron) |
| Gemini 3 Flash | $0.50 / $3.00 | $0.15 / $0.90 (Kie.ai) | $0.20 / $1.20 (Infron) |
| Gemini 2.5 Pro | $1.25 / $10.00 | $0.38 / $3.00 (Kie.ai) | – |
| Gemini 2.5 Flash | $0.30 / $2.50 | $0.09 / $0.75 (Kie.ai) | – |
| Gemini 2.5 Flash-Lite | $0.10 / $0.40 | $0.08 / $0.32 (Infron) | – |
What to know before switching:
- Kie.ai charges about 30% of Google’s price on the Flash and Pro models it carries, and it is the cheapest real-time option wherever it lists a model. It is a reseller, so its terms and rate limits apply, not Google’s.
- Infron sells real-time Gemini through Google Vertex at 20% below Google’s list price, and its Flex routes cost half of that again but are slower. Infron adds 5% + $0.35 to each credit top-up. Its 3.1 Pro and 3.5 Flash-Lite Flex prices are lower than Google’s own Flex rates ($1 / $6 and $0.15 / $1.25).
- Runcrate sells Gemini 2.5 Pro at $0.656 / $5.25 and 2.5 Flash at $0.158 / $1.31, about half of Google’s price.
- OpenRouter also offers Gemini on Google’s Flex tier at half price, the same rate as Google’s own Flex. Its real-time Gemini prices match Google’s, so it doesn’t undercut Google.
- Atlas Cloud, DeepInfra and WaveSpeed AI mostly charge Google’s list price. AIMLAPI, Venice AI and Muapi (on 3.1 Pro) charge more.
The full ranking is at cheapest Gemini APIs, and our Kie.ai pricing breakdown explains how its credits work.
How much does Gemini image generation cost?
Image models have no free API tier. Google charges about $0.0336 per 1K image for Gemini 3.1 Flash Lite Image, $0.039 for Gemini 2.5 Flash Image, $0.067 per 1K image for Gemini 3.1 Flash Image (from $0.045 at 0.5K) and $0.134 for Gemini 3 Pro Image. Kie.ai lists lower starting prices on all four, from $0.02 per image, though its prices rise with resolution. Gemini 2.5 Flash Image shuts down on March 15, 2027. See the cheapest image generation APIs.
Which Gemini models are being retired?
- Gemini 2.5 Pro and 2.5 Flash are no longer in AI Studio. From December 1, 2026, requests to them are routed to Gemini 3.8 Flash and Gemini 3.5 Flash-Lite, so you pay those models’ prices.
- Gemini 3.5 Flash and Gemini 3.7 Flash were removed from Google’s price list. Their endpoints route to 3.6 Flash and 3.8 Flash.
- Gemini 2.0 Flash and Gemini 3 Pro Preview are deprecated, replaced by 3.6 Flash and 3.1 Pro Preview.
Bottom line
For most workloads, Gemini 3.8 Flash is the default choice: free to test, $0.75 / $3.75 until the January price rise, $0.375 / $1.875 on Batch or Flex, and $0.225 / $1.125 through Kie.ai. For the lowest bill on simple tasks, Gemini 2.5 Flash-Lite at $0.10 / $0.40 is the cheapest Gemini model, and Gemma 4 costs nothing on Google’s free tier or a few cents per million tokens on third-party hosts. Prices change often, so check the Google provider page for current figures.