Google AI Studio logo

Google AI Studio

api

by google

Free browser based IDE for prototyping with Gemini models, managing API keys, and building production AI applications on the Gemini Developer API

Key features

Prompt Design and Testing
API Key Management
Full Model Access
Code Generation and Export
Multimodal Input Support
Grounding with Google Search
Pricing

From $undefined/undefined

Best For

Developers who want free API access to Gemini models for prototyping, testing, and small production workloads

Verdict

Exceptionally generous free tier: most models offer free input and output tokens, unmatched by OpenAI or Anthropic

What it does

Prompt Design and Testing

Interactive prompt playground for chat, freeform, and structured prompts. Adjust temperature, top K, top P, safety settings, and system instructions in real time with instant model responses.

Learn more

API Key Management

Create, rotate, and manage Gemini API keys directly from the console. One click key generation with project level organization and usage tracking per key.

Full Model Access

Access the complete Gemini model family: Gemini 3.1 Pro Preview, Gemini 3 Pro, Gemini 3 Flash, Gemini 2.5 Pro, Gemini 2.5 Flash, Gemini 2.5 Flash Lite, Gemini 2.0 Flash, Gemma 3, and more. Switch models instantly to compare quality, speed, and cost.

Learn more

Code Generation and Export

Generate production ready API call code from any prompt in Python, JavaScript, Go, Java, C#, and cURL/REST. Copy and integrate directly into your application.

Multimodal Input Support

Upload and process images, audio files, video clips, PDFs, and code files directly in the prompt playground. Test multimodal capabilities before writing a single line of code.

Grounding with Google Search

Enable Google Search grounding to give models access to real time web information. Responses include source citations for verification.

Learn more

Model Tuning (Fine Tuning)

Fine tune Gemini models on your own data directly in the browser. Upload training examples, configure hyperparameters, monitor training progress, and deploy tuned models with unique endpoints.

Learn more

Generous Free Tier

Most models include free input and output tokens with reasonable rate limits. Prototype, test, and run small workloads at zero cost before scaling to paid usage.

Learn more

Image and Video Generation

Generate images with Imagen 4 and videos with Veo 3 and Veo 3.1 directly from the console. Test creative AI capabilities alongside language models.

Text to Speech Generation

Generate natural speech audio with Gemini 2.5 Flash TTS and Gemini 2.5 Pro TTS models. Preview voice output and configure speech parameters in the browser.

Live API for Voice Agents

Build real time voice applications and agents with the Live API. Test native audio input and output with Gemini 2.5 Flash Native Audio directly in AI Studio.

Batch API

Submit large batches of prompts for asynchronous processing at 50% reduced cost compared to standard API pricing. Available on the paid tier.

Pricing

ModelInput / 1M tokensOutput / 1M tokens
Gemini 3.1 Pro PreviewLatest flagship. $4/$18 for prompts >200K tokens. Paid tier only.$2.00$12.00
Gemini 3 Flash PreviewFrontier intelligence at flash speed. Free tier available.$0.50$3.00
Gemini 2.5 ProUp to 1M context. $2.50/$15 for prompts >200K tokens. Free tier available.$1.25$10.00
Gemini 2.5 FlashHybrid reasoning with 1M context. $1 audio input. Free tier available.$0.30$2.50
Gemini 2.5 Flash LiteMost cost effective model. Free tier available.$0.10$0.40
Gemini 2.0 FlashBalanced multimodal model, 1M context. Free tier available.$0.10$0.40
Gemini 2.0 Flash LiteLightweight model for high volume use. Free tier available.$0.07$0.30
Gemma 3 / Gemma 3nOpen source models. Free tier only, no paid tier pricing.FreeFree
Gemini EmbeddingText embeddings. Free tier available.$0.15N/A

Pros & Cons

Pros

  • Exceptionally generous free tier: most models offer free input and output tokens, unmatched by OpenAI or Anthropic
  • Instant API key generation with no credit card required, reducing friction to start building
  • Access to the full Gemini model family in one place, from lightweight Flash Lite to flagship Gemini 3.1 Pro Preview
  • Built in code generation exports production ready snippets in Python, JavaScript, Go, Java, C#, and REST
  • Grounding with Google Search gives models access to the most comprehensive web index for real time information
  • Model tuning (fine tuning) available directly in the browser without additional infrastructure
  • Batch API offers 50% cost reduction for high volume asynchronous workloads
  • Multimodal playground supports uploading images, audio, video, PDFs, and code files for interactive testing
  • Gemini 2.5 Flash Lite at $0.10/$0.40 per 1M tokens is one of the cheapest capable LLM APIs available
  • Open source Gemma models available completely free, with no paid tier required

Cons

  • Free tier content is used to improve Google products, which may raise privacy concerns for sensitive data
  • Some preview models (Gemini 3.1 Pro, Gemini 3 Pro) are not available on the free tier
  • Rate limits on the free tier are significantly lower than paid, which can block rapid prototyping sessions
  • No native prompt versioning or team collaboration features; you need external tools for prompt management at scale
  • Enterprise features require migrating to Vertex AI on Google Cloud, which is a separate platform with its own pricing
  • Gemini models are generally considered a step behind Claude for nuanced writing and behind OpenAI for certain reasoning tasks, depending on the benchmark

How to get started

1

Open Google AI Studio

Navigate to aistudio.google.com and sign in with your Google account. No separate registration or credit card required.

2

Get your API key

Click "Get API key" in the top navigation. Create a new key or select an existing Google Cloud project. Copy the key for use in your applications.

3

Experiment in the prompt playground

Start a new chat prompt or freeform prompt. Select a model (Gemini 2.5 Flash is a good starting point for speed and cost), type your prompt, and see the response instantly. Upload images, audio, or documents to test multimodal capabilities.

4

Adjust model parameters

Fine tune temperature (creativity), top K, top P, max output tokens, and safety settings in the right panel. Enable Google Search grounding if your application needs real time web information.

5

Export code to your application

Click "Get code" to generate production ready API call code in Python, JavaScript, Go, Java, C#, or cURL. Copy the snippet directly into your project.

6

Install the SDK and build

Install the official Google Generative AI SDK for your language (pip install google-genai, npm install @google/genai) and integrate the exported code into your application. The free tier supports real prototyping and small production workloads.

Deep dive

Detailed guides with comparisons, tips, and visuals for each feature.

Get notified about updates

We'll email you when this tool's pricing or features change.

Last updated: 2026-02-21