by GoogleLLM APIs
Per-token access to Google's Gemini models with an ongoing free tier, billed in US dollars through Google AI Studio and Cloud Billing.
Google's Gemini models by the token in US dollars, with an ongoing free tier to start and batch, flex and caching discounts to scale.
The Gemini API sells Google's text, image, video, speech and embedding models by the token, in US dollars, behind one API key from Google AI Studio. An ongoing free tier covers select current models, including Gemini 3.1 Flash-Lite and 3.8 Flash, with no card needed. Paid use starts with a $5 prepay; Batch and Flex (preview) halve token prices, cached input costs 90% less, and paid prompts are not used to improve Google products. Teams that want to prototype for free and then move the same code to paid use get both on one key.
Against 3 other products we reviewed in LLM APIs: what you pay for with each, the five access facts, the plan limits we recorded and each review's bottom line. The note under the table says how each one was checked.
| Metric | Google Gemini API | Mistral AI API | Anthropic API (Claude) | OpenAI API |
|---|---|---|---|---|
| What you're paying for | Easy to adoptGoogle's Gemini models by the token in US dollars, with an ongoing free tier to start and batch, flex and caching discounts to scale. | Worth the priceLow per-token prices on small open-weight models, with EU or US processing you can choose yourself and spending caps per workspace. | Worth the priceClaude models from Haiku 5.5 at $0.10 / $0.50 per 1M tokens up to Fable 5.1, with no training on API data and a BAA you can sign in the console. | Saves engineering timeOne API key for OpenAI's GPT-6 models and hosted tools, priced per token with published batch, flex and caching discounts. |
| Overall score | 5.3 | 7.4Best | 7.2 | 7.1 |
| Free option | Free plan | Free plan | Free trial | –Not measured |
| Card to start | No card needed | No card needed | –Not measured | –Not measured |
| Getting access | Self-serve | Self-serve | Self-serve | Self-serve |
| Starter priceCheapest paid plan | $0.25/1M input tokens | $0.10/1M input tokens | $0.10/1M input tokens | $0.10/1M input tokens |
| Pricing model | Usage-based | Usage-based | Usage-based | Usage-based |
| Pricing page | Public prices | Public prices | Public prices | Public prices |
| How you pay | Prepaid credits ($5 minimum), expire after 12 months | Prepaid pay-as-you-go credits, optional auto recharge | Prepaid credits by card, expire after 1 year; invoicing through sales | Prepaid credits ($5 minimum), expire after 1 year |
| Free option | Ongoing free tier on select models, no card; limits shown in AI Studio | Free mode, no card; allowance not stated publicly | One-time free credits for new users, amount not stated | No free allowance stated |
| Reference scenario (100M in + 20M out tokens/mo) | $55 on gemini-3.1-flash-lite, computed | $12 on ministral-3b-2512, computed | $20 on claude-haiku-5-5, computed | $20 on gpt-6-luna, computed |
| Batch, cache and other discounts | Batch and Flex (preview) 50% off; cached input 90% off | Batch 50% off; cached input 90% off | Batch 50% off; cache hits 90% off on Haiku 5.5 | Batch and Flex 50% off; cached input 90% off |
| Trained on API data | Paid: no. Free tier: may be used to improve Google products | Unless you opt out (Free mode and paid) | No (Commercial Terms) | No, unless you opt in |
| Default retention | Abuse logs for a "limited period", days not stated | 30 rolling days for abuse monitoring | Deleted within 30 days | Abuse logs up to 30 days |
| Zero data retention | Not guaranteed: paid prompts are logged for abuse monitoring (docs point to Gemini Enterprise Agent Platform, formerly Vertex AI) | By approval, pay-as-you-go, stateless endpoints only | By agreement, through sales | By approval, through sales |
| SOC 2 / ISO 27001 | Not stated for this API | Both, stated in the help center | Both, listed for the API | Both, covering the API |
| HIPAA BAA | Not stated for this API | Not stated | Standard BAA signed in the Console, if eligible | Standard BAA accepted in the API Platform, if eligible |
| Data regions | No region choice stated | EU and US endpoints at +10% | US-only inference at 1.1x; otherwise global | 10 storage regions by approval; processing in US, EU, UAE |
| Official SDKs | Python, JavaScript/TypeScript, Go, Java, C# | Python, TypeScript | Python, TypeScript, C#, Go, Java, PHP, Ruby | Python, TypeScript/JavaScript, Java (beta), Go (beta), C#, Ruby |
| Console SSO | Not stated (Google account sign-in) | SAML on the Enterprise plan | Through sales | Custom/Enterprise billing plan |
| Bottom line | From its public pages, the Gemini API pairs an ongoing free tier on current models with a detailed USD price list and published batch, flex and caching discounts. Certifications, zero data retention and data regions sit with Google Cloud's Agent Platform, and a few billing pages disagree. | From its public pages, the Mistral AI API has the lowest reference price in this comparison, EU or US endpoints you can pick yourself and spending caps. Training use is an opt-out, rate limits sit behind sign-in, and ZDR and SSO need approval or the Enterprise plan. | From its public pages, the Anthropic API has a complete per-token price list, discounts that combine and strong data terms, with a self-serve HIPAA BAA. Region choice is US only, zero data retention and SSO go through sales, and a few pages disagree on billing details. | From its public pages, the OpenAI API has a complete per-token price list, published batch, flex and caching discounts, and audit reports that cover the API. Stricter data controls, residency and SSO go through sales, and a few docs pages lag behind the price list. |
Evidence: Google Gemini API: facts checked 2026-10-15 from its public sign-up and pricing pages, no completed hands-on run; Mistral AI API: facts checked 2026-10-15 from its public sign-up and pricing pages, no completed hands-on run; Anthropic API (Claude): facts checked 2026-10-15 from its public sign-up and pricing pages, no completed hands-on run; OpenAI API: facts checked 2026-10-15 from its public sign-up and pricing pages, no completed hands-on run. Prices in USD as listed, excluding tax. "–" means not recorded.
Review from public pages
Written from Google Gemini API's public pages (pricing, security or privacy, and docs) as read on Oct 15, 2026. We did not sign up, and no hands-on run is published yet.
We read the Gemini API pricing page, the models page and model pages, the deprecations page and release notes, the billing, rate limit, logs, zero data retention, Batch API, caching and Flex docs, the SDK and OpenAI compatibility docs and the Gemini API Additional Terms on Oct 15, 2026, plus Google Cloud's compliance scope pages and Google's processor DPA; we did not sign up, create a key or call the API. The docs say AI Studio creates a project and key for a new user and new accounts begin on the Free Tier; a payment method and a $5 prepay are asked for only when you set up billing for the Paid Tier.
The pricing page is detailed: every model has Free and Paid columns, Standard, Batch, Flex and Priority rows, and a line saying whether content is used to improve Google products. Picking the scenario model takes care. The 2.5 models carry the lowest prices but are now limited to users who used them before, so a new project starts on 3.1 or 3.5 Flash-Lite. Gemini 3.1 Flash-Lite is listed as stable, with a May 7, 2027 earliest shutdown date on the deprecations page.
A few points need a question. The billing page promises a refund of prepaid credits when switching to Postpay, while its FAQ says that switch is not supported. The troubleshooting guide calls the 2.5 models deprecated, while the models page says they are not. Paid prompts are logged for abuse detection "for a limited period of time", with no number of days, and free-tier and paid rate limits are shown only inside AI Studio.
The pricing page lists free input and output tokens on select current models, including the scenario model, with no end date and no card. Google AI Studio is free as well.
vs Mistral AI API: Mistral AI API also has a free mode with no card, but its pages do not say which models it includes; OpenAI API states no free allowance.
Batch is 50% off, Flex (preview, best-effort) is 50% off, and cached input on Gemini 3.1 Flash-Lite costs 90% less than standard input.
vs Anthropic API (Claude): Anthropic API (Claude) has the same batch and cache discounts but no published flex-style lower rate.
Usage is drawn from a prepaid balance, auto-reload is optional and takes a monthly auto-charge limit, and projects can have spend caps (marked experimental).
Compared: DeepSeek API states no spend limit; spending is bounded only by the prepaid balance.
The Google GenAI SDK is GA for Python, JavaScript/TypeScript, Go, Java and C#, structured outputs follow a JSON Schema, and the OpenAI libraries work with a changed base URL.
vs Mistral AI API: Mistral AI API's official SDKs are Python and TypeScript only.
Each pricing table says whether content is used to improve Google products. On the Paid Tier it is not, and prompts fall under Google's processor DPA with the terms.
Google Cloud's SOC 2, ISO 27001 and HIPAA lists name Gemini Enterprise Agent Platform, not this API. The ZDR page points to Vertex AI for guaranteed zero data retention, and the terms say data may be cached in any country with Google facilities.
Workaround: The pages point to Gemini on Gemini Enterprise Agent Platform (formerly Vertex AI), a separate Google Cloud offer with its own prices.
Free-tier content may be used to improve Google products and read by human reviewers, outside the EEA, Switzerland and the UK. Paid prompts are logged for abuse detection for a period the terms do not state; Search grounding keeps prompts 30 days.
Workaround: Link a billing account to move a project to the Paid Tier before sending data that should not be used.
Gemini 2.5 Flash-Lite ($0.10 / $0.40) is limited to users who used 2.5 models before, so a new project starts on 3.1 Flash-Lite, where our scenario is $55 a month (computed). Its earliest shutdown date is May 7, 2027.
Workaround: Batch or Flex halves the scenario to $27.50 (computed) for work that can wait.
Credits expire after 12 months and are non-refundable; at a $0 balance every key on the billing account stops with HTTP 402. Tier 1 is capped at $250 a month, and the refund-on-Postpay rule is contradicted by the FAQ.
Workaround: Set up auto-reload with a monthly limit; Tier 2 (a $2,000 cap) opens after $100 paid and 3 days.
The Gemini API bills per 1M input and output tokens in USD, per model and tier (Standard, Batch, Flex, Priority); thinking tokens count as output. The free tier gives free tokens on select models at limits shown in AI Studio. The Paid Tier needs a Cloud Billing account and a prepay of at least $5 (the most you can prepay is $5,000), and credits expire after 12 months. Our scenario on Gemini 3.1 Flash-Lite is 100 × $0.25 + 20 × $1.50 = $55 a month at Standard rates, or $27.50 on Batch or Flex (computed). Costs climb with Gemini 4 Argon ($2 / $10, rising to $4 / $20 on Dec 1, 2026), the 3.8 and 3.6 Flash increase on Jan 1, 2027, Priority at 75-100% more, hourly storage for explicit caches, and Search grounding at $14 per 1,000 requests after 5,000 free a month. Linking a billing account puts you in Tier 1, with a $250 monthly cap. Taxes fall under the Google Cloud Terms.
The Interactions API stores conversation state by default unless store is set to false. File API uploads stay until you delete them or they expire, explicit caches until their TTL, and opt-in logs for a window you set (7, 14, 28 or 55 days); logs saved in datasets have no set retention. Code on the OpenAI-compatible endpoint moves by changing the base URL and key; Google GenAI SDK code needs porting. Prepaid credits are forfeited if you close the billing account.
From its public pages, the Gemini API pairs an ongoing free tier on current models with a detailed USD price list and published batch, flex and caching discounts. Certifications, zero data retention and data regions sit with Google Cloud's Agent Platform, and a few billing pages disagree.
| Limit | As recorded |
|---|---|
| How you pay | Prepaid credits ($5 minimum), expire after 12 months |
| Free option | Ongoing free tier on select models, no card; limits shown in AI Studio |
| Reference scenario (100M in + 20M out tokens/mo) | $55 on gemini-3.1-flash-lite, computed |
| Batch, cache and other discounts | Batch and Flex (preview) 50% off; cached input 90% off |
| Trained on API data | Paid: no. Free tier: may be used to improve Google products |
| Default retention | Abuse logs for a "limited period", days not stated |
| Zero data retention | Not guaranteed: paid prompts are logged for abuse monitoring (docs point to Gemini Enterprise Agent Platform, formerly Vertex AI) |
| SOC 2 / ISO 27001 | Not stated for this API |
| HIPAA BAA | Not stated for this API |
| Data regions | No region choice stated |
| Official SDKs | Python, JavaScript/TypeScript, Go, Java, C# |
| Console SSO | Not stated (Google account sign-in) |
Read from the vendor's public pages on Oct 15, 2026. No hands-on run is published yet. How we check
We have not published a sign-up run for Google Gemini API yet. The access facts above were read from its sign-up and pricing pages on Oct 15, 2026.
We have not published API calls for Google Gemini API yet.
Google Gemini API lists 5 plans. The cheapest paid plan, Gemini 3.1 Flash-Lite, costs $0.25/1M input tokens.
Prices as listed, USD, excluding tax; observed 2026-10-15.
Source: ai.google.dev pricing page. Prices change; the observed date above is when our agent read them.
We read the pricing, model, data-usage, trust and docs pages of three LLM APIs on the same day and scored them on the same six indicators. What 100 million input and 20 million output tokens a month cost, how each one treats your data, and who should pick which.
