Inside the Gemini factory: Accelerating partner solutions

Google Cloud has introduced the Gemini factory, a co-innovation initiative that helps enterprise software partners rapidly build and deploy agentic solutions using standardized connectors such as Model Context Protocol (MCP), Agent-to-Agent (A2A) runtimes, and REST APIs. The program pairs partner teams with Google Cloud specialists in global hubs like San Francisco, London, and Tokyo to ideate, architect, and deliver production-ready integrations in days rather than months.

The initiative supports four architectural primitives—MCP connectors, declarative Skills (.md), Plugins, and AI agents (A2A)—allowing partners to choose the right pattern for their use case. MCP connectors expose external APIs and CRUD operations, while Skills provide domain-specific instructions and workflows. Plugins bundle these into turnkey solutions, and A2A agents enable proprietary algorithms with tamper-proof audit trails.

Developers use Antigravity, an AI-assisted environment, to scaffold connectors, author Skills, and validate integrations with pre-built templates and automated testing. Partners can host Remote MCP Connectors on Cloud Run or their own cloud, while A2A agents run in isolated environments to protect intellectual property and ensure deterministic execution.

The Gemini factory offers funding incentives, including upfront credits and performance-based credits, alongside go-to-market support such as co-marketing assets and launch campaigns. Partner solutions are published directly to Google Cloud Marketplace, where they gain visibility through AI Solution Finder and streamlined enterprise deployment via OAuth, Dynamic Client Registration, and BYOL flows.