What are Google's main AI productivity tools in 2026?

At the centre is Gemini, Google's family of AI models, surfaced through the Gemini app and Gemini for Google Workspace (Gmail, Docs, Sheets, Meet, Vids). Around it sit developer tools — Antigravity, Jules and the Gemini/Antigravity CLI for coding; Stitch, AI Studio and Opal for design and prototyping; and the Agent Development Kit with Vertex AI for building production agents. NotebookLM (now Gemini Notebook) handles source-grounded research.

What is the difference between the Gemini app and Gemini for Workspace?

The Gemini app is a standalone assistant — a chat interface, like ChatGPT or Claude. Gemini for Workspace is the same intelligence embedded inside Gmail, Docs, Sheets, Meet and the rest, so it can draft, summarise and analyse using your own files and email. Most organisations get the most value from the Workspace integration because it works where the work already happens.

Which Gemini model should I use in 2026?

For everyday chat and Workspace tasks, Gemini 3.5 Flash is the fast default. For the hardest reasoning, long-context and multimodal work, Gemini 3.1 Pro is the flagship (available on Google AI Pro and above). Developers pick per-call on the API, where Flash-Lite is cheapest and Pro is most capable. The right choice is the cheapest model that clears your quality bar.

How much do Google's AI tools cost?

The Gemini app has a free tier. Consumer plans run from Google AI Plus at around US$7.99/month to Google AI Pro at US$19.99/month and Google AI Ultra at US$99.99/month. Gemini for Workspace is included in most Business and Enterprise Workspace plans at no extra cost since January 2025. Developer and enterprise tools like AI Studio, ADK and NotebookLM have free tiers, with usage-based or Cloud pricing above them.

Is Google AI training HRDC claimable in Malaysia?

Yes. Training on Google's AI tools — from Gemini in Workspace to building agents with ADK — is claimable under HRD Corp's SBL-KHAS scheme for eligible Malaysian employers when delivered by a registered provider. See our HRDC AI training guide.

Google's AI Productivity Stack in 2026: Gemini to ADK

Google now ships an AI tool for nearly every job — and it is easy to lose track of what does what. Here is the whole stack, decoded, with a plain answer to "which one do I actually use?"

By Warren Leow 2026-06-17 10 min read

Two years ago, "Google's AI" meant one thing: a chatbot. In 2026 it means a couple of dozen products, and even people who work with this daily lose track of which tool does what. Is Antigravity a competitor to Stitch? Does AI Studio overlap with ADK? Where does NotebookLM fit? Fair questions — the naming does not help.

This is the map. One place to see the whole Google AI stack, grouped by who it is actually for, so you can stop guessing and pick the right tool for the job in front of you. We go deep on each cluster in its own guide; this is the overview that ties them together.

The centre of gravity: Gemini

Everything starts with Gemini, Google's family of AI models. Gemini 3 launched in November 2025 and the line has moved quickly since: Gemini 3.5 Flash is the fast, default model across the Gemini app and Search, while Gemini 3.1 Pro is the flagship for the hardest reasoning and long-context, multimodal work. You meet Gemini in two main ways — the standalone Gemini app, and embedded across Google Workspace.

Google AI tools represented as an abstract connected system — One model family, many front doors: the Gemini models sit under nearly every Google AI product, from Workspace to the developer tools.

The whole stack, by who it is for

The cleanest way to make sense of the sprawl is to sort the tools by the person using them, not by the technology. Four groups cover almost everyone.

Everyone

Gemini app & Gemini for Workspace

The assistant most people meet first — in Gmail, Docs, Sheets, Meet and Vids, plus the standalone Gemini app. The everyday productivity layer, powered by Gemini 3.

Builders & devs

Antigravity, Jules & Gemini CLI

Google's agentic development platform, the async coding agent Jules, and the terminal agent now folding into Antigravity CLI. For people who ship software.

Makers & designers

Stitch, AI Studio & Opal

Turn a prompt into a UI with Stitch, prototype against the Gemini API in AI Studio, and assemble no-code AI mini-apps in Opal. Idea to working screen in one sitting.

Enterprise & engineers

ADK, Vertex AI & NotebookLM

The Agent Development Kit and Vertex AI Agent Builder for production agents; NotebookLM (now Gemini Notebook) for source-grounded research. Where serious workloads live.

For everyone: Gemini in Workspace

If you live in Gmail, Docs, Sheets and Meet, Gemini for Google Workspace is the tool you will touch most — drafting email, summarising threads, building formulas, generating decks and taking meeting notes, grounded in your own content. Since January 2025 it has been included in most Business and Enterprise Workspace plans at no extra cost, which quietly made it the most widely deployed AI assistant in the world. It is Google's direct answer to Microsoft Copilot; we compare the two approaches in our Copilot ecosystem guide. Full breakdown: Gemini for Google Workspace.

For builders: Antigravity, Jules and the CLI

Google's coding story consolidated in 2026 around Antigravity, its agentic development platform, with the terminal-based Gemini CLI folding into the new Antigravity CLI. Alongside sits Jules, an asynchronous agent that clones your repo, plans, runs tests and opens a pull request while you do something else. This is the layer for people who ship software — and the natural home for vibe coding. Deep dive: Antigravity, Jules and Gemini CLI.

For makers: Stitch, AI Studio and Opal

Between "I have an idea" and "I have a working screen", Google now sits three tools. Stitch turns a natural-language prompt into a high-fidelity UI with clean HTML and CSS. AI Studio is the playground for prototyping against the Gemini API. Opal lets you assemble no-code AI mini-apps. Together they collapse design-to-code into a single sitting. See Stitch, AI Studio and Opal.

For engineers and enterprises: ADK, Vertex AI and NotebookLM

When the goal is a production agent rather than a prototype, the tools change. The Agent Development Kit (ADK) is Google's open-source, code-first framework for building agents; Vertex AI Agent Builder — now part of the rebranded Gemini Enterprise Agent Platform — is where they get deployed, scaled and governed. And NotebookLM, recently renamed Gemini Notebook, is the source-grounded research assistant that turns your own documents into summaries, mind maps and audio overviews. See building agents with ADK and Vertex AI and NotebookLM for research.

How to choose, in one line

Use the Gemini app or Workspace to get work done; Antigravity, Jules and the CLI to build software; Stitch, AI Studio and Opal to prototype fast; ADK and Vertex AI to ship production agents; and NotebookLM to make sense of your own sources. That is 90% of the decision.

The part that decides your return

Here is the pattern I keep seeing in Malaysian organisations: the tools are free or nearly free, and the licences get switched on, but usage stays flat because nobody was taught how to work this way. Google keeps shipping capability; the constraint is capability on your side. That gap is a training problem, and it is HRD Corp SBL-KHAS claimable — our AI Vibe Coding, AI Automation and AI Engineering programmes are built to close it, tool-agnostic and mapped to your stack. The funding path is in the HRDC AI training guide.

Put Google's AI tools to work for your team

Knowing the tools is one thing; building with them is another. AITraining2U runs hands-on, HRD Corp SBL-KHAS claimable AI training for Malaysian organisations — from vibe coding and automation to production AI engineering — tool-agnostic and mapped to your stack. Talk to us about a programme for your team.