Gemini Flash Pricing Doubles on January 1: What It Will Cost You
Key Takeaways Gemini API pricing is about to change twice. Gemini 4 Argon doubles on December 1, 2026, and Gemini 3.8 Flash and Gemini 3.6 Flash double on January 1, 2027. Flash costs $0.75 per million input tokens and $3.75 per million output tokens today, and $1.50 and $7.50 from January 1. The Flash-Lite text […]
Gemini API pricing changes twice in five weeks. Gemini 4 Argon doubles on December 1, 2026, and Gemini 3.8 Flash and 3.6 Flash double on January 1, 2027.
Flash goes from $0.75 and $3.75 to $1.50 and $7.50 per million tokens. On a 200M input and 40M output workload, that is $300 a month today and $600 from January, an extra $3,600 a year.
Every paid tier doubles, including caching. Batch, Flex, Priority, cached input and cache storage all move by the same 2x, so caching absorbs none of the increase.
Batch cancels the increase entirely. From January 1, Flash on Batch costs $0.75 and $3.75, exactly what Standard costs today. Argon on Batch works the same way from December 1.
Flash-Lite keeps its current rates. Gemini 3.5 Flash-Lite at $0.30 and $2.50 carries no change date and costs 27% of what Flash will, on the same API and SDK.
Gemini API pricing is about to change twice. Gemini 4 Argon doubles on December 1, 2026, and Gemini 3.8 Flash and Gemini 3.6 Flash double on January 1, 2027. Flash costs $0.75 per million input tokens and $3.75 per million output tokens today, and $1.50 and $7.50 from January 1.
The Flash-Lite text models keep their current rates. Every paid tier of the affected models doubles, including Batch, Flex, Priority and context caching.
This guide covers what every Gemini model costs, what each tier adds, what the increases do to your bill and how to cut your costs before they take effect.
Quick answer
| Question | Answer |
| What does Gemini 3.8 Flash cost? | $0.75 input and $3.75 output per 1M tokens, rising to $1.50 and $7.50 on January 1, 2027 |
| What does Gemini 4 Argon cost? | $2.00 input and $10.00 output per 1M tokens, rising to $4.00 and $20.00 on December 1, 2026 |
| Which models do not change? | Gemini 3.5 Flash-Lite, 3.1 Flash-Lite and 2.5 Flash-Lite carry no price-change date |
| Is there a free tier? | Yes, with limited model access. Gemini 4 Argon is not included |
| Cheapest way to avoid the increase? | Move non-urgent work to Batch or Flex, which will cost what Standard costs today |
How much does the Gemini API cost in 2026?
Paid Gemini API pricing ranges from $0.10 per million input tokens for Gemini 2.5 Flash-Lite to $4.00 for Gemini 4 Argon after December 1. Output costs more than input on every model, and thinking tokens are billed at the output rate.
These are Google’s published Standard rates for the Gemini Developer API, per million tokens.
| Model | Input now | Output now | Input after change | Output after change | Change date |
| Gemini 4 Argon | $2.00 | $10.00 | $4.00 | $20.00 | Dec 1, 2026 |
| Gemini 3.8 Flash | $0.75 | $3.75 | $1.50 | $7.50 | Jan 1, 2027 |
| Gemini 3.6 Flash | $0.75 | $3.75 | $1.50 | $7.50 | Jan 1, 2027 |
| Gemini 3.5 Flash-Lite | $0.30 | $2.50 | $0.30 | $2.50 | No date listed |
| Gemini 3.1 Flash-Lite | $0.25 | $1.50 | $0.25 | $1.50 | No date listed |
| Gemini 2.5 Flash-Lite | $0.10 | $0.40 | $0.10 | $0.40 | No date listed |
A few points the table does not show:
- Gemini 3.1 Pro Preview costs $2.00 input and $12.00 output for prompts up to 200,000 tokens, and $4.00 and $18.00 above that.
- Flash pricing is flat. Gemini 3.8 Flash and 3.6 Flash charge the same rate at every prompt length.
- Gemini 2.5 Flash-Lite is now limited to projects that have used it before, so new projects cannot adopt it.
Upcoming Gemini API pricing changes
Two Gemini API pricing changes are coming. Gemini 4 Argon doubles on December 1, 2026, covered in the next section, and every paid rate for Gemini 3.8 Flash and Gemini 3.6 Flash doubles on January 1, 2027. Flash Standard goes from $0.75 to $1.50 per million input tokens and from $3.75 to $7.50 per million output tokens, and Google shows both prices side by side on its pricing page.
| Tier (per 1M tokens) | Through Dec 31, 2026 | From Jan 1, 2027 |
| Standard input | $0.75 | $1.50 |
| Standard output, including thinking | $3.75 | $7.50 |
| Standard cached input | $0.075 | $0.15 |
| Batch input | $0.375 | $0.75 |
| Batch output | $1.875 | $3.75 |
| Batch cached input | $0.0375 | $0.075 |
| Flex input | $0.375 | $0.75 |
| Flex output | $1.875 | $3.75 |
| Priority input | $1.35 | $2.70 |
| Priority output | $6.75 | $13.50 |
| Priority cached input | $0.135 | $0.27 |
| Cache storage, per 1M tokens per hour | $0.50 | $1.00 |
Five things to know before you budget:
Gemini 4 Argon pricing
Gemini 4 Argon pricing is $2.00 per million input tokens and $10.00 per million output tokens through November 30, 2026. From December 1, 2026 it is $4.00 and $20.00. Argon is Google’s flagship reasoning model, and its increase lands a month before the Flash change.
| Tier (per 1M tokens) | Through Nov 30, 2026 | From Dec 1, 2026 |
| Standard input | $2.00 | $4.00 |
| Standard output | $10.00 | $20.00 |
| Batch and Flex input | $1.00 | $2.00 |
| Batch and Flex output | $5.00 | $10.00 |
Gemini 4 Argon is not available on the free tier.
The Batch move works here too. From December 1, Argon on Batch costs $2.00 and $10.00, exactly what Standard costs today.
If you run both Argon and Flash, your Gemini bill rises twice, 31 days apart.
Your bill before and after the increases
A workload of 200 million input tokens and 40 million output tokens a month costs $300 on Gemini 3.8 Flash today and $600 from January 1. That is an extra $3,600 a year on one model, with no change to your traffic, prompts or code.
| Model and date | Input cost | Output cost | Monthly | Annual |
| Gemini 3.8 Flash, today | $150 | $150 | $300 | $3,600 |
| Gemini 3.8 Flash, from Jan 1 | $300 | $300 | $600 | $7,200 |
| Gemini 4 Argon, today | $400 | $400 | $800 | $9,600 |
| Gemini 4 Argon, from Dec 1 | $800 | $800 | $1,600 | $19,200 |
The same workload on Argon goes from $800 to $1,600 a month, an extra $9,600 a year.
On a monthly invoice, a rate change looks like growth: spend doubles while traffic stays flat. Track cost per million tokens as well as the total, so you can tell which one moved. Our LLM cost calculator runs your own token counts against both rate cards.
Gemini API free, paid and enterprise tiers
The Gemini API has a free tier, but it limits which models you can use, and Google uses your content to improve its products. The Paid tier removes those limits and adds Batch pricing and context caching. Enterprise customers buy through Gemini Enterprise Agent Platform.
| Tier | What you get |
| Free | Limited access to certain models, free input and output tokens, Google AI Studio access. Content is used to improve Google’s products |
| Paid | Access to Google’s most advanced models, higher rate limits, context caching, Batch API at a 50% reduction. Content is not used to improve Google’s products |
| Enterprise (Gemini Enterprise Agent Platform) | Everything in Paid, plus optional dedicated support, advanced security and compliance, provisioned throughput and volume-based discounts. Sold through Google’s sales team |
Rate limits on the Paid tier rise as your billing history with Google grows. Google shows each model’s current limits in the Google AI Studio console rather than in a fixed table, so check there before you plan capacity.
The free tier suits prototypes and testing. Production traffic belongs on Paid, both for the rate limits and because free-tier content is used for product improvement.
Gemini Batch, Flex, Priority and caching prices
Batch and Flex cost half of Standard, and Priority costs 80% more. Cached input costs a tenth of the Standard input rate, plus a storage fee per million tokens per hour. On Gemini 3.8 Flash and 3.6 Flash, all of these double on January 1, 2027.
| Option (Gemini 3.8 Flash) | Input now | Output now | Compared with Standard |
| Standard | $0.75 | $3.75 | Baseline |
| Batch | $0.375 | $1.875 | 50% less |
| Flex | $0.375 | $1.875 | 50% less |
| Priority | $1.35 | $6.75 | 80% more |
| Cached input (Standard) | $0.075 | n/a | 90% less on input |
How to use each one:
- Batch is for work that does not need an instant answer, such as overnight enrichment, bulk classification, summarization backlogs and evaluation runs.
- Flex costs the same as Batch on both sides of the January date.
- Priority costs more per token. Use it only for traffic where the higher price is justified.
- Context caching pays off when many requests reuse the same long prompt or document. Add the storage fee to your estimate: $0.50 per million tokens per hour now, $1.00 from January.
Gemini grounding, image and audio prices
Grounding with Google Search includes 5,000 free requests a month, shared across Gemini 3 and newer models, then costs $14 per 1,000 requests. Image and audio output are priced per million output tokens, at much higher rates than text.
| Feature | Price | Changes on Jan 1, 2027? |
| Grounding with Google Search | 5,000 free requests a month, then $14 per 1,000 | No |
| Nano Banana 2.1 image output | $30 per 1M output tokens, about $0.034 per 1K-resolution image | No date listed |
| Gemini 3.8 Flash TTS audio output | $9.00 per 1M tokens | Yes, to $18.00 |
| Gemini 3.8 Flash-Lite TTS text input | $0.50 per 1M tokens | Yes, to $1.00 |
| Gemini 3.8 Flash-Lite TTS audio output | $6.00 per 1M tokens | Yes, to $12.00 |
If your app grounds answers in search results, the grounding fee can matter as much as token costs once you pass the free 5,000 requests. Count grounded requests separately when you estimate a monthly bill.
Gemini on Agent Platform (Vertex AI) vs the Gemini API
Pricing matches at global endpoints. Gemini Enterprise Agent Platform, formerly Vertex AI, lists the same January 1, 2027 change for Gemini 3.8 Flash and 3.6 Flash: $0.75 and $3.75 now, $1.50 and $7.50 from January 1. Regional (non-global) endpoints cost 10% more.
| Endpoint | Through Dec 31, 2026 | From Jan 1, 2027 |
| Global | $0.75 / $3.75 | $1.50 / $7.50 |
| Regional (non-global) | $0.825 / $4.125 | $1.65 / $8.25 |
Differences that matter on Agent Platform:
- The current price works as a credit. Google applies the introductory pricing as 50% credits back on net spend, so check your invoice credits, not just list prices.
- Gemini 3.6 Flash retires earlier. On Agent Platform it retires on November 19, 2026, before the price change, and Gemini 3.8 Flash replaces it.
- Gemini 3.7 Flash is still listed on Agent Platform with the same doubling, and it retires on January 28, 2027.
- Gemini 2.5 Flash-Lite has already retired on Agent Platform. Google lists 3.8 Flash, 3.5 Flash-Lite or Gemma 4 as replacements.
- Negotiated deals are separate. Provisioned throughput and volume discounts are agreed with Google’s sales team, so if you have one, ask your account team how the new rates apply.
Agent Platform usage appears with the rest of your Google Cloud costs, so review it next to your compute and storage. For other Google Cloud service rates, see our GCP pricing catalog.
How to calculate your Gemini API cost
Multiply your monthly input tokens and output tokens by the rate for each, then add caching, storage and any grounding or media charges. Use the rate in effect on the date the tokens are billed, since Argon changes on December 1 and Flash on January 1.
Follow these steps:
Worked example: Gemini 3.8 Flash from January 1, with 200 million input tokens (50 million of them from the cache) and 40 million output tokens.
| Line | Calculation | Cost |
| Uncached input | 150M × $1.50 | $225.00 |
| Cached input | 50M × $0.15 | $7.50 |
| Output | 40M × $7.50 | $300.00 |
| Total before storage | $532.50 |
Add cache storage based on how many tokens you keep cached and for how many hours. You can run the same numbers on your own traffic with our free calculator.
How to cut Gemini API costs before the price increases
Moving asynchronous work to Batch or Flex cancels the increase. From January 1, Flash on Batch costs $0.75 and $3.75, exactly what Standard costs today. After that, the biggest savings come from moving to a smaller model.
Move non-urgent work to Batch or Flex. This means overnight jobs, bulk classification, evaluations and backlogs. You keep the same model with no quality trade-off.
Test Gemini 3.5 Flash-Lite. Google describes it as a cost-efficient model optimized for high-volume agentic tasks, translation and simple data processing, and names it as one of two models to use for new projects. It stays on the same API and SDK.
Compare other providers at their full rates. The table below prices the 200M input / 40M output workload.
Run evals on your own traffic. Find the cheapest model that meets your quality bar, and price it at the new rates rather than today’s.
Watch cost per million tokens after each change. The first week of December and of January will show whether the change hit as expected.
| Model | Input per 1M | Output per 1M | Monthly cost |
| Gemini 3.8 Flash, from Jan 1 | $1.50 | $7.50 | $600 |
| Gemini 3.5 Flash-Lite | $0.30 | $2.50 | $160 |
| Gemini 3.1 Flash-Lite | $0.25 | $1.50 | $110 |
| OpenAI GPT-6 Luna | $0.10 | $0.50 | $40 |
| Claude Haiku 5.5 (prompts up to 100K tokens) | $0.10 | $0.50 | $40 |
On this workload, Gemini 3.5 Flash-Lite costs 27% of what Flash will. Gemini 3.1 Flash-Lite is cheaper, but Google lists May 7, 2027 as its earliest shutdown date, with 3.5 Flash-Lite as the replacement.
Claude Haiku 5.5 rises to $0.50 and $2.50 for prompts above 100,000 tokens. Token counts also differ between providers, so measure them on your own traffic. Switching providers means a second vendor and a migration, so weigh that against the saving. See our full guides to OpenAI API pricing and Claude API pricing.
How Economize tracks Gemini and AI API costs
Economize shows your AI API costs next to your AWS, Azure and GCP spend, with cost per million tokens and spend by model. That lets you tell a price change from usage growth on the day it happens, not at month end.
- Cost per 1M tokens, not just the total. If the rate doubles, your total doubles and so does cost per token. If your usage doubles, the total doubles while cost per token holds. The invoice alone cannot tell you which happened.
- Spend by model. Each model has its own row, so a move from Flash to Flash-Lite shows up as a shift in model mix you can verify.
- Anomaly detection with root cause attached. This matters most in the first week of December and January, when a price step and a usage spike look the same in a monthly view.
- AI spend beside your cloud bill. Economize covers Vertex AI (Gemini Enterprise Agent Platform), OpenAI, Anthropic, AWS Bedrock, Azure AI Services and Mistral, all in the same place as your cloud costs.
- Published, flat pricing. Free up to $100,000 a month in tracked spend, with no percentage-of-spend fees.
Setup is agentless and takes about five minutes.
Frequently Asked Questions
Gemini API pricing runs from $0.10 per million input tokens (Gemini 2.5 Flash-Lite) to $2.00 (Gemini 4 Argon), rising to $4.00 for Argon on December 1, 2026. Gemini 3.8 Flash costs $0.75 input and $3.75 output today. Economize tracks what you actually spend per model.
There is a free tier with limited model access, and Google uses free-tier content to improve its products. Gemini 4 Argon is not on the free tier. Production use needs the Paid tier.
Twice. Gemini 4 Argon doubles on December 1, 2026, from $2.00 and $10.00 to $4.00 and $20.00 per million tokens. Gemini 3.8 Flash and 3.6 Flash double on January 1, 2027. Economize’s anomaly detection flags the step change in your spend when each one lands.
Gemini 3.8 Flash and 3.6 Flash cost $1.50 per million input tokens and $7.50 per million output tokens from January 1, 2027. Batch and Flex cost $0.75 and $3.75, which is what Standard costs today, and Priority costs $2.70 and $13.50.
No. Gemini 3.5 Flash-Lite ($0.30 / 2.50),3.1Flash-Lite(0.25 / 1.50)and2.5Flash-Lite(0.10 / $0.40) carry no price-change date. The Gemini 3.8 Flash-Lite TTS voice model does double on January 1.
Yes. Gemini Enterprise Agent Platform, formerly Vertex AI, lists the same January 1, 2027 change: $0.75 and $3.75 rising to $1.50 and $7.50 at global endpoints, with regional endpoints 10% higher. Economize shows Vertex AI spend next to the rest of your Google Cloud bill.
Connect your Google Cloud billing to Economize to see Vertex AI spend by model, cost per million tokens and anomalies, next to your AWS, Azure, OpenAI and Anthropic costs. See how Economize customers track AI and cloud spend in one place.
Plan for both price increases now
A price change you know about is a budget line. One you miss is a surprise on next month’s invoice. Gemini 4 Argon doubles on December 1 and Gemini Flash on January 1, so now is the time to move batchable work, test Flash-Lite and set up tracking.
If cloud cost management is new to your team, FinOps 101 covers the foundations, and our cloud cost optimization guide covers the wider practice. Connect your accounts to Economize to track cost per token and spend by model for free, with no credit card needed.
Simran Sardar
FinOps enthusiastProduct Manager at Economize with over 3 years of experience, focused on FinOps strategies and cloud cost optimization. Dedicated to helping organizations streamline cloud expenses and drive financial efficiency.
Maximize Cloud Efficiency and Optimize Costs
Get started free in our sandbox or book a personalized call with our experts
