Google Introduces Tabletop, a Collaborative AI Planning Experiment
Google Labs has launched Tabletop, an experimental platform that allows groups to build customized planning tools using conversational AI agents.
Google LLC has initiated testing for Tabletop, a fresh experiment from Google Labs designed to help groups create custom planning tools with the assistance of an artificial intelligence agent. According to Roman Nurik, the design and engineering lead at Google Labs, coordinating shared projects like family vacations or community meals is easier when participants collaborate within a single online hub.
Collaborative AI Tool Generation
When initiating a Tabletop space, the creator uploads gathered materials such as links, documents, and reference ideas. Users then interact with an @tabletop handle to generate the specific instrument they require. Google highlights sample outputs including structured itineraries, dinner selection polls, expense distribution calculators, and tournament brackets, alongside a collection of initial templates.
Invitations are distributed via shared links dropped into existing communication channels, permitting simultaneous collaboration akin to traditional shared document editing. For instance, demonstration scenarios show how a request can transform an office party seating chart into an interactive layout where name badges can be repositioned between tables. Any generated utility remains fully adjustable through ongoing conversational prompts.
While Google did not specify the exact underlying model powering Tabletop, prior releases like the Disco browser experiment utilized Gemini technology to convert open browser tabs into micro-applications. This release follows related collaborative rollouts, such as Google's September initiative enabling household groups to share the CC agent, and OpenAI's addition of group chat capabilities for up to 20 users within ChatGPT threads.
What it means for developers
As collaborative, multi-user AI agent environments become standard across consumer and enterprise applications, developers must build systems capable of handling shared state, real-time synchronization, and multi-participant prompting. Building these kinds of collaborative AI workflows often requires experimenting with multiple frontier models to test response quality, latency, and agent behavior in group contexts. Developers can try top AI models cheaply through one API at https://apixoai.online.
Availability and Access
Currently, Tabletop is restricted to users in the United States who are 18 years of age or older. Interested participants can join a waitlist via tabletop.google, with Google stating that access will expand incrementally on a weekly basis. Official details regarding pricing or international availability outside the United States have not yet been disclosed.
Source: Google tests Tabletop, letting groups build planning tools with an AI agent — SiliconANGLE AI. Written by the Apixo team from that report.
One key for Claude, GPT, GLM, DeepSeek and more. Pay per token with crypto.
Get your API key
