Google ADK
Google ADK (Agent Development Kit) is an agent development toolkit released by Google to help developers build, orchestrate, and deploy AI agent applications based on large language models. The toolkit provides core capabilities including agent definition, tool integration, multi-agent collaboration, and session management. It supports connectivity with Google Cloud and third-party services and is suited for building complex automated task-processing and conversational AI systems.
Core Facts
Timeline (last 90 days)
Google ADK has the broadest language support among the six frameworks in this article, offering four SDKs for Python, TypeScript, Java, and Go, and natively supporting the A2A protocol.
Google ADK (Agent Development Kit) targets the Gemini ecosystem and provides multi-agent collaboration and deployment support.
Strands itself does not provide search tools and requires connecting to an MCP server; ADK has built-in google_search but only for Gemini models.
Strands uses MCPClient to connect directly to the AWS documentation MCP server, while ADK integrates it through McpToolset.
In terms of page reading, Strands' web_fetch supports markdown mode to return the original text, making it easy to verify facts word by word; ADK's load_web_page has problems with redirects and short-line filtering, causing large numbers of code examples to be lost.
ADK sets output_schema= on the Agent, and the result is stored in the session state as a dictionary, requiring a manual call to model_validate for conversion.
When an ADK Agent has both an output_schema and built-in tools, ADK automatically adds a function tool named set_model_response, causing a conflict with Gemini API built-in tools, and include_server_side_tool_invocations=True must be set to resolve it.
ADK nodes can be any function, which is more flexible, but there is no built-in loop count limit at all, and the routing function must decide on its own when to stop.
Strands adds a hook to each Agent via add_hook to check cumulative cost before model calls; ADK achieves similar functionality via before_model_callback and after_model_callback, and supports a plugin mechanism so that one registration applies to all Agents.
Strands wraps exceptions into EventLoopException after tool execution; ADK first emits an error event and then throws the exception at the end of the stream.
10 more timeline events
All Facts (20)
Google ADK (Agent Development Kit) is an agent development framework launched by Google.
80%Unverified2024-2025年，Anthropic Claude Agent SDK、OpenAI Agents SDK、Google ADK等Agent框架集中涌现
85%UnverifiedGoogle ADK not only supports Google's own Gemini model, but can also be adapted to other models.
60%UnverifiedGoogle ADK has the broadest language support among the six frameworks in this article, offering four SDKs for Python, TypeScript, Java, and Go, and natively supporting the A2A protocol.
50%UnverifiedGoogle ADK (Agent Development Kit) targets the Gemini ecosystem and provides multi-agent collaboration and deployment support.
50%UnverifiedStrands wraps exceptions into EventLoopException after tool execution; ADK first emits an error event and then throws the exception at the end of the stream.
50%UnverifiedStrands itself does not provide search tools and requires connecting to an MCP server; ADK has built-in google_search but only for Gemini models.
50%UnverifiedStrands uses MCPClient to connect directly to the AWS documentation MCP server, while ADK integrates it through McpToolset.
50%UnverifiedIn terms of page reading, Strands' web_fetch supports markdown mode to return the original text, making it easy to verify facts word by word; ADK's load_web_page has problems with redirects and short-line filtering, causing large numbers of code examples to be lost.
50%UnverifiedADK sets output_schema= on the Agent, and the result is stored in the session state as a dictionary, requiring a manual call to model_validate for conversion.
50%UnverifiedWhen an ADK Agent has both an output_schema and built-in tools, ADK automatically adds a function tool named set_model_response, causing a conflict with Gemini API built-in tools, and include_server_side_tool_invocations=True must be set to resolve it.
50%UnverifiedADK nodes can be any function, which is more flexible, but there is no built-in loop count limit at all, and the routing function must decide on its own when to stop.
50%UnverifiedStrands adds a hook to each Agent via add_hook to check cumulative cost before model calls; ADK achieves similar functionality via before_model_callback and after_model_callback, and supports a plugin mechanism so that one registration applies to all Agents.
50%UnverifiedA survey of 12 mainstream Agent frameworks found that none of the frameworks provides monetary-denominated tool-call budget control in its documentation (C5 is all no).
50%UnverifiedPexafy's official documentation lists tested configurations for the OpenAI Agents SDK, Claude Agent SDK, Google ADK, CrewAI, LlamaIndex, Mastra and Strands Agents.
50%UnverifiedPink Agentic AI Payments conducted a systematic review of the 12 most popular Agent frameworks and concluded that none of the frameworks documents a currency-denominated tool-call budget control (C5).
50%UnverifiedThe frameworks covered by the scanner include LangChain, LangGraph, LlamaIndex, CrewAI, AutoGPT, Flowise, n8n, Google ADK, and Semantic Kernel.
50%UnverifiedAgent framework selection is fundamentally an engineering constraints problem: choose MAF for the Microsoft stack, Google ADK for GCP, Mastra for TS teams, LlamaIndex for data processing, CrewAI for role mapping, OpenAI Agents SDK for minimalist low-level needs, and LangChain for cross-model compatibility.
50%UnverifiedWhen Google ADK is deployed on Cloud Run, container restarts cause in-memory state to be lost, and external persistent storage must be configured before going live.
50%UnverifiedGoogle ADK is a GCP-native framework that works out of the box and comes with built-in debugging capabilities.
50%