Subvolts take Google's Gemini API gets a developer boost with new JavaScript examples and advanced features. Oct 15, 2026 - Developers
- You gain updated JavaScript examples and tools to build more capable Gemini-powered applications with new features.
- Business
- Your teams can leverage enhanced Gemini capabilities like Search grounding and context caching for better AI integrations.
Developers can now use updated JavaScript examples to leverage Gemini's latest features, including Search grounding and Gemini 3.
Gemini API Cookbook: feat(quickstarts-js): migrate Get_Started.js to Interactions API (#1413)
feat(quickstarts-js): migrate Get_Started.js to Interactions API (#1413) * feat(quickstarts-js): align Get_Started.js with Python quickstarts - Add Thinking controls (includeThoughts and thinkingBudget) - Remove deprecated sampling parameters (temperature, topP, topK) - Add Google Search and Google Maps grounding - Add Context Caching workflow (create, list, query, delete) - Add Text and Multimodal Embeddings with gemini-embedding-2 - Add Gemini 3 section (thinking levels, media resolution, thought signatures, migration guide) - Modernize model selector with gemini-3.8-flash default * fix(quickstarts-js): safely access candidate parts with optional chaining * refactor(quickstarts-js): migrate Get_Started.js to Interactions API * Address review comments for Get_Started.js * docs(quickstarts-js): clean up unused imports and update resolution docs in Get_Started.js
Opening of the original on Gemini API Cookbook From Subvolts
Summary and analysis by Subvolts editorial, AI-assisted.
Google updated its Gemini API Cookbook with a JavaScript quickstart migration to the Interactions API. This update aligns the JavaScript example with Python versions, incorporating features like thinking controls, removing deprecated sampling parameters, and adding Google Search and Maps grounding. It also introduces context caching, text and multimodal embeddings, and details on Gemini 3 capabilities, including thinking levels and media resolution.
Google updated its coding guide for Gemini. Developers can now use newer, simpler code to build apps that understand context better and can even use Google Search and Maps. It also shows off Gemini 3's smarter thinking and ability to handle more complex media.
This update focuses on developer tools for Gemini. By migrating the JavaScript quickstart to the Interactions API, Google makes it easier for developers to build applications using the latest Gemini features. The inclusion of Search and Maps grounding, context caching, and embeddings provides powerful new capabilities. Developers can now more easily experiment with Gemini 3's advanced features like thinking levels and higher media resolution, enhancing multimodal application development.
- Updated Quickstart: Get_Started.js migrated to Interactions API
- New Features: Thinking controls, Google Search/Maps grounding, Context Caching, Embeddings
- Removed Parameters: Deprecated sampling parameters (temperature, topP, topK)
- Default Model: gemini-3.8-flash
- Embedding Model: gemini-embedding-2
This update enhances Gemini's developer tools, mirroring efforts by OpenAI and Anthropic to simplify API integration and add new model capabilities.
- Interactions API
- An API designed to streamline common conversational AI tasks and manage model interactions.
- A new way to talk to Gemini that makes building apps easier.
- Sampling Parameters
- Parameters like temperature, topP, and topK that control the randomness and diversity of model output.
- Settings that control how creative or predictable the AI's answers are.
- Grounding
- Connecting a language model's output to external, verifiable information sources like search engines.
- Making sure the AI's answers are based on real information from places like Google Search.
- Context Caching
- Storing previous conversation turns to maintain state and improve efficiency in long dialogues.
- Remembering past parts of a conversation so the AI doesn't lose track.
Explore the updated JavaScript quickstart in the Gemini API Cookbook. Try running the code to see how it integrates Google Search results or caches conversational context.
The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news. From the Playbook, matched to the product and topics of this item.
- ToolAstrocade platform for text-based game creationBuild and customize personalized interactive games using natural language prompts without coding. · 30 min
- PromptDemonstrate AI progress by comparing past and present capabilitiesShow how far AI has come since a specific past AI capability by having the AI perform a task. · 5 min
- PromptCreate alphabet video with unusual items and lower thirdsGenerate a rapid-fire alphabet video with unique items and text labels. · 15 min
- PromptGenerate claymation explainer of protein foldingCreate a stop-motion claymation video explaining protein folding accurately. · 10 min
Questions people ask
- What is the Interactions API?
- The Interactions API is a newer way to interact with Gemini models, simplifying common tasks and enabling advanced features like context caching.
- What are the new Gemini 3 features mentioned?
- Gemini 3 features include adjustable thinking levels, higher media resolution, and thought signatures, with a migration guide provided.
More from Gemini API Cookbook 85 more
Page generated Oct 19, 2026. Summaries are Subvolts' own; the story belongs to Gemini API Cookbook.