Gemini API billing changes in March and April 2026: spend caps by tier ($250 to $100K+ a month, enforced from April 1), prepay for new users and, from a cutover reported as October 12, 2026, for existing Postpay accounts, and Gemini 3.1 Pro Preview paid only. The free tier covers the Gemini 3.x Flash and Flash-Lite models.
Published 2026-03-26 · Reviewed 2026-10-21 · Spend caps enforced April 1 · 3.1 Pro paid-only · Prepaid billing live · Figures compiled 2026-03-26, last checked 2026-10-21 · 52 of 52 figures in the “Free LLM API Comparison” table come from our records for 1,493 developer tools
$250
Tier 1 Spend Cap
Dec 2025
Free tier cut
0
Free Pro models for new projects
8
Providers Compared
Google began enforcing monthly spend caps on the Gemini API on April 1, 2026 (Tier 1: $250, Tier 2: $2,000, Tier 3: $20,000 to $100,000+). When a billing account reaches its cap, requests pause until the next billing month. New users default to Prepay (minimum $5), and existing Postpay accounts must switch to Prepay by the cutover date in their account notice, reported as October 12, 2026. Gemini 3.1 Pro Preview is paid only.
For developers who built on Gemini's generous early free tier, the API has fundamentally changed: On 2025-12-06 Google cut 2.5 Flash's free tier from 250 requests a day to about 20, and 2.5 Pro's to none. The free tier covers the Gemini 3.x Flash and Flash-Lite models. Google publishes no free-tier limits; AI Studio shows each project's. Since 2026-09-18 Google serves the Gemini 2.5 models only to users who used them before. New projects use 3.5 Flash-Lite or 3.8 Flash. Below we cover what changed, who's affected, and which alternatives offer better free access.
From our tracker: Google cut the Gemini API free tier's rate limits on 2025-12-06. Gemini 2.5 Flash went from 10 requests a minute and 250 a day to 5 and 20, and Gemini 2.5 Pro went to no free requests. Google's rate-limits page stopped listing free-tier limits the same day. Source ↗
Google Gemini API — Free tier: Gemini 3.8 Flash, 3.6 Flash, 3.5 Flash-Lite, 3.1 Flash-Lite, Gemini 3 Flash Preview, Gemini Embedding 2 and Gemma 4 are free of charge. Also free of charge are Gemini 3.8 Live (real-time voice), Gemini 3.5 Live Translate (translation), Gemini 3.5 Transcribe (speech-to-text), Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS (text-to-speech). Gemini 4 Argon (gemini-4-argon), Gemini 3.1 Pro Preview and Gemini Omni Flash (gemini-omni-1.1-flash) are paid only. Gemini 4 Argon, generally available since 2026-10-13, is not available on the Gemini API free tier. According to ai.google.dev/gemini-api/docs/google-ai-plans, "Gemini 4 Argon is not available on the Free plan." Gemini 4 Argon is available through the Gemini API on the paid tier. "Direct use of the Gemini API (such as using API keys or external applications) is billed and managed separately," per ai.google.dev/gemini-api/docs/google-ai-plans. Google AI Pro and Ultra include Gemini 4 Argon in Google AI Studio and Antigravity (Pro: non-trial subscriptions only), per ai.google.dev/gemini-api/docs/google-ai-plans and antigravity.google/docs/models. No Gemini image model has a free tier in the API. Nano Banana 2.1 (gemini-nano-banana-2.1) became generally available on 2026-10-06. It costs $0.0336 per 1K image, $0.0504 per 2K image and $0.113 per 4K image. The Batch API halves those prices. Nano Banana Pro costs $0.134 per 1K or 2K image and $0.24 per 4K image. Google publishes no free-tier rate limits; each project's limits are shown in Google AI Studio. Developers report 429 errors showing 20 requests per day for Gemini 3.x Flash models and 500 requests per day for 3.x Flash-Lite models as of September 2026. Rate limits apply per project, not per API key. Daily request quotas reset at midnight Pacific time, per Google's rate-limits page. Since 2026-09-18 Google serves the Gemini 2.5 models only to users who have used them before, and points new projects to 3.5 Flash-Lite or 3.8 Flash. If your app calls gemini-3.7-flash, Google deprecated it on 2026-10-08. Requests to gemini-3.7-flash are routed automatically to gemini-3.8-flash. Google says to update the model string to gemini-3.8-flash. If your app calls Gemini 3.5 Flash (gemini-3.5-flash), Google deprecated it on 2026-10-08. Requests to gemini-3.5-flash are routed automatically to gemini-3.6-flash. Google says to update the model string to gemini-3.6-flash or gemini-3.8-flash. If your app calls gemini-omni-flash-preview, it can be called until 2026-10-22, when Google shuts it down. Google's replacement is gemini-omni-1.1-flash. Gemini 3.1 Flash-Lite (gemini-3.1-flash-lite) can be called until 2027-05-07, when Google shuts it down. Google's replacement is gemini-3.5-flash-lite. Paid, per million tokens (input/output; output prices include thinking tokens): Gemini 4 Argon $2/$10 through 2026-11-30 (context caching $0.10, then $0.20, implicit caching only); Gemini 3.8 Flash $0.75/$3.75 until 2026-12-31, then $1.50 input and $7.50 output from 2027-01-01 (3.6 Flash: the same prices and dates); Gemini 3.5 Flash-Lite $0.30/$2.50; Gemini 3.1 Flash-Lite $0.25/$1.50; Gemini 3.1 Pro Preview $2/$12 (prompts up to 200K tokens; prompts over 200K tokens cost $4 input and $18 output). From December 1, 2026, Gemini 4 Argon costs $4 per million input tokens and $20 per million output tokens for Standard, $2 input and $10 output for Batch API, $2 input and $10 output for Flex, and $7.20 input and $36 output for Priority. The Batch API costs half the standard price; Gemini 3.8 Flash is $0.375 input and $1.875 output per million tokens through 2026-12-31. Gemini 4 Argon on the Batch API is $1 input and $5 output through 2026-11-30. Google also sells two tiers, Flex and Priority, both in Preview. A request picks one with the service_tier field ("flex" or "priority"); without it the request is Standard. Gemini 4 Argon on Flex is $1 input and $5 output through 2026-11-30; on Priority it is $3.60 input and $18 output through 2026-11-30. "The Gemini Flex API is an inference tier that offers a 50% cost reduction compared to standard rates, in exchange for variable latency and best-effort availability" (ai.google.dev/gemini-api/docs/flex-inference). "Priority inference is priced at 75-100% more than the standard API and billed per token" (ai.google.dev/gemini-api/docs/priority-inference). The Gemini API Contributor Tier is an invite-only private preview. Eligible users will be invited by email. It offers half the Standard price for Gemini 4 Argon, Gemini 3.1 Pro Preview, Gemini 3.8 Flash, and Gemini 3.5 Flash-Lite. Google uses the data from those models to improve its products. "Please don't submit sensitive or personal information" (ai.google.dev/gemini-api/docs/logs-policy). Grounding with Google Search is supported on the free tier for Gemini 3.8 Live, Gemini 3.8 Live Extended Thinking and Gemini 3.1 Flash Live Preview. It is not available on the free tier for Gemini 4 Argon or the Gemini 3.x Flash, Flash-Lite and Pro models. On the paid tier, 5,000 search requests a month are free, shared across Gemini 3 and newer models, then $14 per 1,000. Accounts opened after 2026-03-02 cannot spend the $300 Google Cloud welcome credit on the Gemini API.
For developers outside the European Economic Area, Switzerland and the United Kingdom, Google's terms say human reviewers may read free-tier prompts and responses, and Google says not to send sensitive, confidential or personal information on the free tier. (From ai.google.dev/gemini-api/terms, read 2026-10-04.)
Apps that serve users in the European Economic Area, Switzerland or the United Kingdom may use only the paid tier. (From ai.google.dev/gemini-api/terms, read 2026-10-04.)
Google's terms require users to be 18 years of age or older to use the Gemini API and Google AI Studio. The API may not be used in a website, app, or other service directed towards or likely to be accessed by individuals under 18. These rules apply to both the free and paid tiers. (From ai.google.dev/gemini-api/terms, read 2026-10-17.)
Google offers the Gemini API and Google AI Studio, free and paid, in 230 countries and territories, including Taiwan, Brazil, Japan, Singapore, and the United Kingdom. China, Hong Kong, Macao, Russia, Belarus, and Iran are not among them. Google says to developers outside the list: "try the Gemini API in Gemini Enterprise Agent Platform". (From ai.google.dev/gemini-api/docs/available-regions, read 2026-10-17.)
Google's terms say changes to Gemini API paid prices take effect 30 days after Google posts them unless Google specifies otherwise. Pricing for a new paid service takes effect immediately unless Google specifies otherwise. (From ai.google.dev/gemini-api/terms, read 2026-10-17.)
Google is moving existing Postpay accounts to Prepay for Gemini API usage. An account that has not switched and added credits by the cutover date in its account notice has its paid service interrupted; other Google Cloud services on the billing account stay on Postpay. (From ai.google.dev/gemini-api/docs/billing, read 2026-10-07.)
When the Prepay balance reaches $0, every API key in every project on the billing account stops working until credits are added; requests then fail with an HTTP 402 Payment Required error. (From ai.google.dev/gemini-api/docs/billing, read 2026-10-07.)
Google's steps for switching to Prepay tell you to set up auto-reload on the AI Studio Billing page to top up your credit balance when it runs low. (From ai.google.dev/gemini-api/docs/billing, read 2026-10-07.)
The monthly Google Cloud credits from Google AI Pro or Ultra are eligible Google Cloud credits. On the Prepay billing plan, these credits pay for Gemini API usage only after you buy Prepay credits in AI Studio. With a $0 Prepay balance, requests fail even with unused Google Cloud credits, according to a Google staff member on discuss.ai.google.dev. (From discuss.ai.google.dev/t/error-code-402-message-your-prepayment-credits-are-depleted/185524, read 2026-10-10.)
A paid project does not automatically revert to the free tier, even when its Prepay balance reaches $0. To use the free tier, disable billing on the project, which unlinks it from its billing account. (From ai.google.dev/gemini-api/docs/billing, read 2026-10-10.)
Disabling billing on a Google Cloud project stops all other paid Google Cloud services in it. Some resources may be deleted and become non-recoverable. Outstanding charges are still due. (From cloud.google.com/billing/docs/how-to/modify-project, read 2026-10-10.)
We read that on 2026-10-03 from https://ai.google.dev/pricing, and our check recorded: the page names Google Gemini API as "gemini api" and states "$0.75" and "$1.50" and "$3.75" and "$7.50". A source check reads the cited page for the service's name and a price; the limits above are from our own record.
Not published. About 5 RPM and 20 RPD on 2.5 Flash, none on 2.5 Pro (2025-12-06)
92% fewer daily requests on 2.5 Flash
Gemini 2.0 Flash
Available, standard limits
Shut down 2026-06-01
Google recommends 3.6 Flash or 3.1 Flash-Lite
Spend caps (Apr 1)
No hard caps — billed without pausing
$250/mo (Tier 1), $2K/mo (Tier 2), $20K+ (Tier 3)
Requests pause at cap
Billing model
Pay-as-you-go for all
Prepay by default for new users; Postpay accounts must switch (reported cutover October 12, 2026)
Minimum $5 prepayment
Gemini 3.1 Pro
N/A (new model)
Paid only
No free tier
Context window
1M tokens
1M tokens
Unchanged
Flash-Lite limits
Generous (unspecified)
Not published
Free tier preserved
Spend cap details: Google began enforcing billing-account-level spend caps on April 1, 2026. When a tier's spend cap is reached, API requests pause until the next billing month. Affects pay-as-you-go developers who may see unexpected request pausing. Source ↗ Developers on pay-as-you-go plans should set project-level budget alerts in Google Cloud Console to avoid unexpected pausing.
2. Timeline of Gemini API Changes
Dated changes to the Gemini API's free tier and billing.
Dec 2025
Free Tier Cut
Google cut the Gemini API free tier's rate limits on 2025-12-06. Gemini 2.5 Flash went from 10 requests a minute and 250 a day to 5 and 20, and Gemini 2.5 Pro went to no free requests. Google's rate-limits page stopped listing free-tier limits the same day. Source ↗
Mar 2026
$300 Welcome Credit Excluded
Accounts opened after 2026-03-02 cannot spend the $300 Google Cloud welcome credit on the Gemini API or AI Studio. Source ↗
Apr 2026
Spend Caps Enforced
Google began enforcing billing-account-level spend caps on April 1, 2026. When a tier's spend cap is reached, API requests pause until the next billing month. Affects pay-as-you-go developers who may see unexpected request pausing. Source ↗
Sep 2026
2.5 Models Limited to Existing Users
Since 2026-09-18 Google serves the Gemini 2.5 models only to users who have used them before, and points new projects to 3.5 Flash-Lite or 3.8 Flash. Source ↗
Jan 2027 (scheduled)
Flash Prices Double
Gemini 3.8 Flash and 3.6 Flash cost $0.75/$3.75 per million input/output tokens through 2026-12-31 and $1.50/$7.50 from 2027-01-01, by Google's pricing page. Source ↗
3. Who's Affected
The impact depends on how you use the Gemini API and when you started building on it.
Free tier developers (most affected)
If you built on the free tier before December 2025: on 2025-12-06, 2.5 Flash's daily requests fell from 250 to about 20 and 2.5 Pro's to none. Gemini 3.1 Pro Preview has no free tier. Consider switching to Groq (30 RPM free) or OpenRouter (25+ free models).
Pay-as-you-go developers
The April 1 spend caps affect you directly. If your billing account hits its tier's spend cap, requests pause until the next billing month — not just rate-limited, but fully paused. Set project-level budget alerts now. Teams sharing a billing account are especially vulnerable since all projects count toward the same cap.
Low-volume / hobby developers
If you make a few requests a minute on a 3.x Flash model, the free tier still works — just with less headroom. The 1M token context window remains Gemini's key differentiator. For occasional use, the impact is minimal.
4. New Tier Structure & Spend Caps
Google revamped the Gemini API billing structure with automatic tier upgrades based on spend level. Each tier has a monthly spend cap that pauses all API requests when reached.
Tier
Monthly Spend Cap
Free Rate Limits
Paid Rate Limits
Key Constraint
Free
$0
Not published; shown per project in AI Studio
N/A
No free Pro model for new projects.
Tier 1 (Pay-as-you-go)
$250/mo
Same as free
Higher RPM per model
Starts when you link an active billing account. Requests pause at $250 aggregate spend
Tier 2
$2,000/mo
Same as free
Even higher RPM
Automatic once $100 has been paid and 3 days have passed since the first successful payment
Tier 3+
$20K–$100K+
Same as free
Highest RPM, priority
Automatic once $1,000 has been paid and 30 days have passed since the first successful payment
What "spend cap" means in practice: Unlike rate limits (which reject individual requests), spend caps pause all requests for the remainder of the billing month once the tier's aggregate spend limit is reached. This is a billing-account-level control — all projects under the same billing account share the cap. Google added project-level spend caps on March 12, 2026.
Spend-based rate limit: paid tiers can also be limited to a maximum spend over any rolling 10 minutes: $10 on Tier 1, $50 on Tier 2 and $200 on Tier 3. Above it, the API returns a 429 RESOURCE_EXHAUSTED error. Google says whether this limit applies depends on the account's billing history. The Tier 2 and Tier 3 payment thresholds count all Google Cloud spending on the billing account, not only the Gemini API.
5. Prepaid Billing & Paid-Only Models
Two additional changes that affect paid and high-usage developers.
Prepaid billing for all paid accounts
New users default to Prepay and buy at least $5 of credits to set up billing. To keep a high-spend account running, you can prepay up to $5,000 in credits. For amounts larger than $5,000 a month, use auto-reload to automatically add credits. If you set a monthly auto-reload limit, auto-reload stops when reached. Manual purchases still work. Auto-reload resumes at the next billing cycle start. If you start the switch to Prepay and stop before paying, other projects may lose service. To restore service, complete the setup, or contact Cloud Billing Support to set the account back to Postpay. Google is moving existing paid accounts from Postpay to Prepay for Gemini API usage: switch on the AI Studio Billing page and buy credits before the cutover date in your account notice, reported as October 12, 2026, or the account's paid Gemini API service is interrupted. An account that missed its cutover date can still switch. Google's billing FAQ says an existing Postpay account can transition to Prepay on the AI Studio Billing page. The doc gives no deadline for the switch. Paid Gemini API service resumes only after the credit purchase is confirmed. Only Gemini API usage moves to Prepay; other Google Cloud services on the same billing account stay on Postpay. Being charged monthly does not exempt an account. Self-serve accounts, charged automatically to a card or bank account, move to Prepay. Prepay is not available for invoiced (offline) accounts, which a business applies for (requirements include one year registered and $40,000 a year of expected spend) and pays by check or wire transfer; Google does not say how the cutover treats them, so holders of one who received a cutover notice should ask Cloud Billing Support. On any account, Gemini Enterprise Agent Platform (formerly Vertex AI) bills after use on the standard Cloud charging cycle. If your Cloud Billing account has a Gemini API Prepay credit balance but your card expired or a payment for another Google Cloud service was declined, the Gemini API can still be suspended. This happens with an overdue balance, a declined payment, or an invalid payment method. Resolve the issue in the Google Cloud Billing console to restore service. Google's billing FAQ states that accounts cannot switch from Prepay to Postpay. To pay after use instead of prepaying, use Gemini Enterprise Agent Platform (formerly Vertex AI). Google bills it on the standard Cloud charging cycle, not Prepay. Calls to gemini-3.5-flash cost $1.50 input and $9.00 output per million tokens on Agent Platform, compared to $0.75 input and $3.75 output on the Gemini API, which routes to gemini-3.6-flash. Change the model ID to gemini-3.8-flash before moving; it costs the same on both platforms. Agent Platform retired gemini-2.5-pro, gemini-2.5-flash and gemini-2.5-flash-lite on October 20, 2026. gemini-3.6-flash can be called on Agent Platform until November 19, 2026, when Google retires it there. On the Gemini API, none of these four models has a shutdown date. Agent Platform charges $0.75 input and $3.75 output per million tokens for Gemini 3.8 Flash and 3.6 Flash on the global endpoint through 2026-12-31, then $1.50 input and $7.50 output. Non-global endpoints cost 10% more. These match Gemini API paid-tier prices for gemini-3.6-flash and gemini-3.8-flash. Accounts that use only the free tier need take no action. Eligible Google Cloud credits are used only after you have bought Prepay credits, and stop being used when the Prepay balance reaches $0. Prepaid credits expire 12 months after purchase and are non-refundable. Prepay credits remain with the Cloud Billing account they were purchased on, so a project moving to another account will not use them. Google staff state funds cannot be transferred between billing accounts, and closing the account forfeits any remaining credits. When the balance reaches $0, every API key on the billing account stops working until you add credits. Calls failing with HTTP 403 "Your project has been denied access" are not from a $0 Prepay balance, which returns 402. Google staff say 403 or "Unavailable" in AI Studio means a flag is on the account. For free-tier projects, complete age verification in your Google Account, then retry or create a new API key. If access is still denied, staff suggest adding Prepay credits, which ends your free tier quota. If you pay by bank transfer, some payments take several days to clear. Most card payments are instant. Gemini API service resumes only after the purchase is confirmed. Usage can continue for approximately 10 minutes past a $0 balance due to billing latency. Batch jobs and agents may consume credits beyond the balance. A negative balance is deducted from the next credit purchase. After buying credits, calls may still fail with "Your prepayment credits are depleted". Google staff say there is no manual billing sync to request. Check that the credits are a Prepay balance on the AI Studio Billing page, not a spend cap or funds in the Google Cloud billing account, and that the API key belongs to a project on that billing account. Retry the calls; if the error continues, post the exact error message and request time on Google's AI developer forum. The AI Studio Billing page may show "The billing account type is inactive or unsupported". Google's billing FAQ says this means the billing account's type or status is not eligible for the Paid Tier. Check the Cloud Console for the account's status. For a Free Trial account, activate billing in AI Studio. For a Closed account, reopen it. There is no manual billing sync to request.
Gemini 3.1 Pro is paid-only
Gemini 3.1 Pro Preview has no free tier. The free tier covers the 3.x Flash and Flash-Lite models, including 3.8 Flash.
Flash-Lite models remain free
The Gemini 3.x Flash and Flash-Lite models are free. Google publishes no free-tier limits; AI Studio shows each project's. For lightweight tasks such as classification, extraction and simple Q&A, 3.5 Flash-Lite or 3.1 Flash-Lite costs nothing.
Claude API with usage-based pricing per million tokens (input/output). Claude Fable 5.1 $10/$50. Claude Opus 5.5 $4/$20. Claude Sonnet 5.5 $2/$10. Claude Sonnet 5 $2/$10. Full profile
AI API platform. One model is priced Free in OpenAI's own table: the moderation model omni-moderation-latest. No GPT model is priced free. Full profile
$0.10/$0.50 (gpt-6-luna) – $10.00/$50.00 (gpt-6-astra) per MTok
Fast LLM inference on Groq's LPU hardware. Free plan: gpt-oss-120b, gpt-oss-20b and Qwen3.8 27B, each at 30 requests a minute, 1,000 requests and 200,000 tokens a day, plus Whisper speech-to-text at 2,000 requests a day. Full profile
$0.075/$0.30 (gpt-oss-20b), $0.15/$0.60 (gpt-oss-120b), $0.80/$4.00 (Qwen3.8 27B) per MTok
free tier caution
Rate limits apply to the whole organization, not to each user or API key. A request over a limit gets a 429 Too Many Requests error. (From console.groq.com/docs/rate-limits, read 2026-10-05.)
You are billed only after you add a payment method and move to the paid Developer tier. (From console.groq.com/docs/billing-faqs, read 2026-10-05.)
Groq's Services Agreement does not permit it to train or fine-tune models on your inputs or outputs unless you give permission. (From console.groq.com/docs/legal/services-agreement, read 2026-10-05.)
Groq does not keep inference requests by default. It may log inputs and outputs for up to 30 days to troubleshoot or investigate abuse; enabling Zero Data Retention in Data Controls stops this. (From console.groq.com/docs/your-data, read 2026-10-05.)
Data Groq retains is stored in Google Cloud buckets in the United States. (From console.groq.com/docs/your-data, read 2026-10-05.)
Mistral's Free plan includes monthly usage shared across Studio, the API and Vibe Code. The pricing page no longer states the amount as of 2026-10-07; it previously listed $10 a month in API credits. Full profile
$0.1/$0.1 (Ministral 3) – $1.5/$7.5 (Mistral Medium 3.5) per MTok
free tier caution
Since August 2026, Mistral offers Mistral AI Studio and its API only to business customers. Its consumer terms for personal use exclude both, and API use falls under the Commercial Terms for an organization, company or other entity. (From legal.mistral.ai/terms/row-consumer-terms, read 2026-10-11.)
A business can use Mistral's API in a commercial product on any plan, including the Free plan; individuals cannot use the API for personal projects. (From legal.mistral.ai/terms/commercial-terms-of-service, read 2026-10-11.)
Models whose names start with "labs-" are free of charge; Mistral calls them experimental and not for production. (From docs.mistral.ai/inference/model-lifecycle, read 2026-10-11.)
Mistral may train on inputs and outputs from Labs or Preview Models, which its Commercial Terms define as experimental, preview or pre-release models identified as such, including those with the "labs" prefix, even if you opted out or use zero data retention. Mistral Large 4 is a public preview. (From legal.mistral.ai/terms/commercial-terms-of-service, read 2026-10-11.)
AI model router. Free plan: 50 requests a day, no BYOK. Free models are capped at 20 requests a minute; accounts that have bought at least $10 of credits get 1,000 free-model requests a day. Full profile
Ultra-fast LLM inference API — no permanent free tier. New accounts get $5 in free credits after adding a verified payment method, expiring 30 days after they are granted. Full profile
$0.35/$0.75 (GPT OSS 120B), $0.99/$1.49 (Qwen 3.8 27B) per MTok
free tier removed
Cerebras's pricing page puts its $5 trial credit and pay-as-you-go pricing on its Developer tier, which it says is not intended for production use, and puts production workloads on its Enterprise tier, priced by quote. (From cerebras.ai/pricing, read 2026-10-09.)
Cerebras's terms give it no right to use API inputs and outputs to train or fine-tune models. (From cerebras.ai/terms-of-service, read 2026-10-04.)
Cerebras's privacy policy says it does not retain inference inputs and outputs. (From cerebras.ai/privacy-policy, read 2026-10-04.)
Free Trial context windows are 65K tokens for gpt-oss-120b and 64K for qwen-3.8-27b. Paid accounts get 131K and 128K. (From inference-docs.cerebras.ai/models/overview, read 2026-10-11.)
Pay-as-you-go LLM API with no free tier, billed from a prepaid balance. DeepSeek-V4.1-Flash (deepseek-flash): $0.30/M input, $1.20/M output at peak. DeepSeek-V4-Pro (deepseek-v4-pro): $1.32/M input, $3.96/M output at peak. Full profile
$0.30/$1.20 (DeepSeek-V4.1-Flash), $1.32/$3.96 (DeepSeek-V4-Pro) per MTok
—
Until November 1, 2026, 00:00 Beijing time, server-side Search and Fetch calls via DeepSeek Harness, Claude Code, or Codex are free; after that, they are billed at $0.75 per 1,000 successful requests. (From api-docs.deepseek.com/quick_start/pricing, read 2026-10-17.)
DeepSeek's Search API and Fetch API are standalone APIs billed per request. Search finds web pages for a query; Fetch retrieves one web page as Markdown. Each costs $0.75 per 1,000 successful requests. Failed requests, empty search results, and timeouts are not billed. (From api-docs.deepseek.com/guides/search, read 2026-10-17.)
Key takeaway:Groq's free plan allows 30 requests a minute and 1,000 a day per model. Mistral AI's Free plan includes monthly API usage; its pricing page listed $10 a month until 2026-10-07, then stopped stating the amount. OpenRouter serves 25+ free models through one API. For the full comparison, see our Free LLM APIs guide.
7. What to Do
Practical steps depending on your situation, from monitoring spend to migrating to alternatives.
If you're staying on Gemini
1. Set project-level spend caps in AI Studio (available since March 12, 2026; Google marks them experimental). This prevents one project from consuming the entire billing account's spend cap. 2. Monitor usage via the Gemini API usage dashboard. Set billing alerts before you hit the spend cap. 3. Move to the 3.x models — Gemini 2.0 Flash shut down on 2026-06-01, and new projects cannot use the 2.5 models. Google points new projects to gemini-3.5-flash-lite or gemini-3.8-flash. 4. Separate billing accounts if you run multiple projects — every project on a billing account shares its tier spend cap.
If you're evaluating alternatives
1. For maximum free requests:Groq — 30 RPM, no credit card, ultra-fast inference. 2. For model variety:OpenRouter — 25+ free models through one OpenAI-compatible API. 3. For long context: Gemini's Flash models accept up to 1,048,576 input tokens. Google no longer publishes free-tier token limits, so check your project's limits in AI Studio before relying on long prompts at no cost. 4. For production workloads:Anthropic and OpenAI also cap monthly spend by usage tier. Anthropic pauses API usage at its tier's cap ($500 a month on Start) until the next month, and OpenAI sets each organization a monthly usage limit ($500 on Build, reached at $5 in total credit purchases).
Bottom Line
If you need a free LLM API today:
Groq and OpenRouter publish their free limits, which Google no longer does: Groq's free plan allows 30 requests a minute and 1,000 a day per model; OpenRouter's free models allow 20 a minute and 50 a day.
If you need the 1M context window:
Stay on Gemini Flash. Set budget alerts, and use 3.5 Flash-Lite or 3.8 Flash for new work.
If you're a paying Gemini customer:
Set project-level spend caps in AI Studio (available since March 12, 2026). Separate billing accounts for production vs. development. Monitor the billing documentation for tier details.
Related Guides
More resources for evaluating LLM APIs and tracking pricing changes.
Methodology: Gemini API pricing details sourced from Google's Gemini API pricing page and billing documentation. Rate limit reductions recorded in our deal change tracker (3 Gemini changes tracked since December 2025). Compiled 2026-03-26, last checked 2026-10-21. Risk assessments are based on pricing history and free tier stability.
This guide covers Gemini API pricing changes through September 2026. For the full LLM API comparison, see Free LLM APIs. For Gemini alternatives, see /alternative-to/google-gemini-api. Browse all 1,493 developer tools at /search.
More Alternatives Guides
LocalStack CE Alternatives— In March 2026, LocalStack merged Community and Pro images, requiring an auth token for the latest image — compare 9 free open-source AWS emulators
Postman Alternatives— Postman killed free team collaboration March 1, 2026 — 5 free API testing alternatives
HCP Terraform Alternatives— HCP Terraform legacy free plan ended March 31, 2026; free tier now caps managed resources at 500 — free IaC alternatives compared
Freshping Alternatives— Freshping shut down March 6, 2026 — 13 free uptime monitoring alternatives
Heroku Alternatives— Heroku removed free tier Nov 2022, entered sustaining mode Feb 2026 — 8 free PaaS options
Firebase Alternatives— Firebase Studio is closing (no new workspaces since June 22, 2026; shutdown March 22, 2027) + Cloud Storage for Firebase now requires Blaze — 7 BaaS alternatives
GitHub Actions Alternatives— GitHub postponed its self-hosted runner fee, so self-hosted runners stay free — 10 free CI/CD alternatives compared
Best Free AI & ML Tools for Developers in 2026— 65+ AI/ML tools and their free tiers compared — LLM APIs, AI coding assistants, ML platforms, observability, and specialized AI services
Best Free LLM APIs in 2026— 25+ LLM API providers and their free tiers compared — proprietary model APIs, open-model inference platforms, and AI gateways with exact rate limits
Best Free API Development Tools in 2026— 39+ free API development tools compared — REST/GraphQL clients, mocking, documentation, marketplaces, and integration platforms
The Complete Free Startup Stack for 2026— Complete free SaaS infrastructure stack — 10 categories with recommended picks, scaling guidance, and stability ratings
The Complete Free AI/ML Stack for 2026— Complete free AI/ML development stack — 10 categories with recommended picks, scaling guidance, and stability ratings
The Complete Free DevOps Stack for 2026— Complete free DevOps infrastructure stack — 10 categories with recommended picks, scaling guidance, and stability ratings
The Complete Free Frontend Stack for 2026— Complete free frontend/Jamstack development stack — 10 categories with recommended picks, scaling guidance, and stability ratings
The Complete Free Next.js Stack for 2026— Complete free Next.js full-stack infrastructure — 10 layers with recommended picks, growth cost analysis, and stability ratings
Google Developer Program 2026— Standalone Google Developer Program Premium no longer takes sign-ups — current plans, Cloud credits and free alternatives
Vercel vs Netlify Free Tier Comparison— Deep comparison of Vercel and Netlify free tiers — bandwidth, functions, builds, commercial use, and scaling costs
Railway vs Render Free Tier Comparison— Deep comparison of Railway and Render free tiers — usage-based vs fixed pricing, databases, sleep behavior, and scaling costs
Datadog vs New Relic Free Tier Comparison— Deep comparison of Datadog and New Relic free tiers — per-host vs per-GB pricing, APM, logs, synthetics, and scaling costs
Free Tier Risk Index— Predictive risk analysis for developer free tiers — grades dated and scored against what happened next, category heatmap, pattern analysis, counter-trends
Free Tier Tracker 2026— A list of developer tool free tiers that were removed or cut in 2026, grouped by quarter, with dates and what changed.
Startup Credits Comparison 2026— The definitive startup credits comparison — 15 programs across cloud infrastructure, fintech, and developer tools with eligibility requirements, vesting schedules, and stacking strategies
AI Coding Tools Pricing Comparison 2026— The definitive AI coding tools comparison — 17 tools across IDE, CLI, cloud agent, and app builder categories with free tier analysis and cost breakdowns
CI/CD Tools Pricing Comparison 2026— The definitive CI/CD pricing comparison — 17+ tools across general, cloud-native, mobile, and self-hosted categories with free tier analysis and cost breakdowns
Database Pricing Comparison 2026— The definitive database pricing comparison — 25+ services across managed Postgres, serverless/edge, document/NoSQL, cloud provider, and specialized categories with free tier analysis and cost breakdowns
Vector Database Pricing Comparison 2026— The definitive vector database pricing comparison — 11 services across dedicated cloud, open-source, pgvector, embedded, and serverless categories with free tier analysis for RAG/AI
Cloud Hosting & PaaS Pricing Comparison 2026— The definitive cloud hosting pricing comparison — 15 platforms across PaaS, edge/serverless, full-featured, and static categories with free tier analysis, pricing gotchas, and Railway referral
LLM API Free Tiers & Free Credits 2026— Which LLM APIs have a genuinely free tier or free credits — frontier labs, inference providers, open-source hosts, and specialized services with free tier analysis and token cost breakdowns
AWS Free Tier 2026— AWS Free plan: up to $200 in credits over 6 months, 30+ always-free services, short-term trials, hidden costs, cheaper alternatives.
GCP Free Tier Complete Guide 2026— Complete GCP free tier guide — 20+ always-free products, $300 trial, hidden costs, and comparison with AWS and Azure
Azure Free Tier 2026— Azure free tier guide — 65+ always-free services, $200 trial, Cosmos DB lifetime free tier, and comparison with AWS and GCP
Testing & QA Tools Free Tier Comparison 2026— Side-by-side comparison of 15+ testing tool free tiers — E2E, visual regression, load testing, API testing, local dev, and the testing cost trap at scale
API Development Tools Free Tier Comparison 2026— Side-by-side comparison of 12+ API development tool free tiers — users, collections, requests, mock servers, local-first vs cloud, and the API tool migration trap
Hosting & PaaS Free Tier Comparison 2026— Side-by-side comparison of 12+ hosting free tiers — bandwidth, compute, build minutes, cold starts, commercial use restrictions, and the hosting cost trap at scale
Developer Security Tools Free Tier Comparison 2026— Side-by-side comparison of 20+ developer security tool free tiers — SAST, SCA, DAST, secrets detection, container security, and the DevSecOps cost trap at scale
Firebase Studio Shutdown Guide— Firebase Studio shuts down March 22, 2027 — free cloud IDE alternatives with compute, storage, and collaboration limits compared
Developer Tool Shutdown Tracker 2026— Living tracker of developer tool shutdowns, API sunsets, and deprecation deadlines in 2026 — with migration paths and alternatives
Forming a US Company in 2026: Costs and Changes— Firstbase Start costs $99 under an offer plus a required $299-a-year agent. Stripe pays the $500 Atlas fee back for a $5,000 Treasury deposit. Delaware's LLC tax is $400.