Updated for Google I/O 2026

Every Google AI tool,
decoded in one tidy place.

Gemini. Antigravity. Flow. Veo. Omni. Spark. NotebookLM. Pics. Stitch. It's a lot. This is a field guide to what each tool actually does and — more usefully — how it differs from the ones it overlaps with.

The three overlaps that trip everyone up

Most "wait, isn't that the same thing?" moments come from these three families. Here's the short version.

🎬 Video: Veo vs Omni vs Flow

Veo 3.1The underlying text-to-video model. Generates clips with audio and dialogue from a prompt.
OmniThe newer multimodal model. Takes any input (image, audio, video, text) and creates or edits video conversationally.
FlowThe creative studio that wraps Veo + Omni + Imagen with reusable characters, scenes, and Flow Tools.

💻 Antigravity: which surface?

Antigravity IDEFull-featured editor with the agent manager, artifacts, and codebase context. Best when you're writing code.
Antigravity 2.0Standalone desktop "mission control" for orchestrating agents in parallel. Use alongside any editor.
Antigravity CLITerminal-first. Replaces the old Gemini CLI. Great for automation, CI, and headless workflows.
Antigravity SDKBuild custom agents on your own infra, or deploy via the Managed Agents API.

🤖 Assistant: Gemini app vs Spark vs Intelligence

Gemini appThe conversational assistant. You ask, it answers, it can run Deep Research and Live voice.
SparkAn agent that acts on your behalf — keeps working in the cloud after you close the app.
IntelligenceAn OS-level layer (Android, ChromeOS, Wear OS) that anticipates intent across third-party apps.

The full catalog

Filter by what you're trying to do, or search by name.

No tools match that yet.