Cross-Language Agent Framework Performance Benchmarking: Google ADK Go vs Python Agent Ecosystems

Faculty-advised research — team forming

→ Core contributors included as co-authors

Agent Framework Research

First systematic performance benchmark across multi-language agent development frameworks, covering latency, throughput, memory, and framework overhead under concurrent and sustained load conditions across Go and Python ecosystems.

Agent FrameworksBenchmarkingGoPythonLangGraphPerformanceMulti-Agent

The crisis

Agent framework performance directly determines deployment cost and latency for AI systems in resource-constrained environments
No standardized benchmark exists for comparing Go vs Python agent frameworks — practitioners make framework choices without empirical data
Framework overhead is invisible in current tooling — organizations optimize LLM calls while ignoring the framework layer beneath them

About this research

As agent development frameworks proliferate across languages and providers, practitioners lack empirical data for comparing them under realistic conditions. This work establishes the first cross-language benchmark covering Google ADK Go, ADK Python, LangGraph, and OpenAI Agents SDK across five scenarios — from single-agent baseline to high-concurrency sustained load. The benchmark isolates framework overhead from LLM API time, enabling fair comparison of the coordination layer itself.

Key findings

(In Progress)

Apply

Join our Slack community to follow updates and see what we're working on.

Apply Now →

← Previous: AI-Mediated Coordination Under Intermediary-Dependency Constraints: Multi-Agent Market Access Across Two Field Deployment Verticals Next: Multi-Agent Specialist Debate for Abstract Visual Concept Learning Across Bongard Problem Taxonomies →