Gemini for Developers - Google Generative AI - Build AI Apps

Learn four ways to access Gemini and other Google models: the Gemini console, Google AI Studio, Vertex AI Studio, and API/SDK with authentication.

Explore Google AI Studio’s playground to chat with Gemini models, adjust temperature and thinking mode, and manage an API key, Google search grounding, and safety settings for tailored prompts.

Access Gemini models from your code using Google AI Studio, generating Python and Java code to prompt models. Use a Gemini API key to enable external access and track costs.

Create an API key by selecting or creating a project in Google AI Studio, then compare free tier and tier one projects for billing implications.

Create and name an API key in Google AI Studio, assign it to a project with a quota tier, then copy the key ID to access a model in Python.

learn to access a Gemini model from Python using the Google Gen AI SDK, including installing the SDK, configuring the API key, and streaming responses.

Create AI apps in Google AI Studio using the Build AI Apps feature with a prompt, upload an image, choose cartoon styles, preview results, download code, and deploy.

Deploy your Tuneify AI app with Google AI Studio and Google Cloud, then access it via Cloud Run, share a URL, and process images or PDFs.

Explore Vertex AI as Google's end-to-end AI platform for Gemini development, from model garden and notebooks to Studio, deployment, and integration with GCP services like GCS, Compute Engine, and BigQuery.

Create a Vertex AI API key, test it with a curl call to Gemini 2.5 Flashlight, and explore service accounts, IAM roles, and key restrictions in the Learn ADK project.

Learn to access a model with a Vertex AI API key in Python, configuring the environment and comparing Vertex AI keys with Google AI Studio keys for GCP projects.

Learn to access a Gemini model using a Google Cloud service account by creating the service account, granting a Vertex AI user role, and generating a JSON key for authentication.

Access a Gemini model with a service account key by configuring project, region, and model name, while following best practices like secret manager and workload identity federation.

Authenticate to Google Cloud using your Google ID for local testing, via application default login, without keys. Grant Vertex AI user permissions and specify your project ID for the program.

Explore Colab Enterprise in Vertex AI to run Python notebooks and access a Gemini model without a service account, API key, or application default login.

Explore Workbench in Vertex AI and create a Workbench instance, selecting region, machine type, and a dedicated service account; compare with Colab Enterprise and leverage GCS and Git integration.

Explore using Google Cloud workbench with JupyterLab to run notebooks and manage VM instances. Access models via federation without service account keys while handling IAM permissions and cost implications.

Install and use the Google GenAI SDK to connect to Gemini and Google models, using a Gemini API key or GCP authentication, and choose models for content or streaming responses.

Limit the thinking budget with GenerateContentConfig and ThinkingBudget, setting zero or a cap (e.g., 300) and balancing model choice (Flashlight, Flash, Pro) to control tokens and cost.

Limit the output tokens with max output tokens and thinking budget to obtain a concise, complete LLM response. Use prompt framing and optional summarization to stay within word limits.

learn how to control randomness in Google Gemini by adjusting temperature, top P, and top K using the generate content config, with recommended ranges and use cases.

Configure safety settings in generate content config to filter outputs for harm, harassment, hate speech, sexually explicit content, and civic integrity. Set thresholds to none, low, medium, or high.

Set a system instruction to assign a persona to the LLM, such as a history expert, and use the generate content class for detailed information with a shared context.

Learn how to stream large language model outputs in Python with generate content stream, printing chunks as the model writes for chat-style apps.

Explore nano banana, a Gemini 2.5 flash image model with image extension, for fast, high-quality image generation. Compare with ImageN for commercial images, and see a cat riding a bike.

Discover editing images with the nano banana multi-modal model, combining image and text prompts to add a red shoe to a cat and save the result.

Generate a five-second video using Google's videos model family, selecting a model such as 3.1 preview, and download the mp4 after rendering.

Learn to summarize a YouTube video with Gemini 2.5 plus by passing a URL or video file, generating a concise three-sentence summary.

Learn grounding for LLMs by connecting to external data sources to overcome knowledge cutoff and enable real-time answers using Google search in the Gemini SDK.

Define a custom schedule meeting function with attendees, date, time, and topic, pass it to the LLM, and explicitly invoke the function to schedule the meeting.

Learn how to install and use the Gemini CLI from the docs quick start, including Node.js and npm requirements, OS-specific install commands, and upcoming exploration.

Authenticate the Gemini CLI with Google ID, test installation by running Gemini, and explore switching to API key or Vertex AI, plus use in cloud shells for development.

Explore the Gemini CLI options, learn how to switch models with the model sub-command, use MCP for local server testing, view stats, and run non-interactive commands from scripts.

Authenticate to the Gemini CLI using a Gemini API key, switch between login with Google and Vertex AI, and manage quotas by choosing API key versus a personal account.

Learn how to authenticate the Gemini CLI with Vertex AI by setting Google Cloud Project and location environment variables, then use Vertex AI for login and logout.

Explore Gemini CLI in Cloud Shell, bundled by default, to create and edit Python files, add documentation, and manage GCP resources like buckets and IAM permissions with minimal prompts.

Explore Gemini Code Assist, a Google plugin that runs beside your code in any major IDE to help you code, debug, fix, and document, with a PyCharm installation walkthrough.

Explore how Gemini code assist helps with Python coding by generating documentation, creating or updating files across multiple directories, and debugging a workspace.