Search papers, labs, and topics across Lattice.
MASEval is introduced as a framework-agnostic library designed to evaluate entire multi-agent systems, rather than focusing solely on the LLM component. It enables systematic comparison of different system implementations by considering factors like topology, orchestration logic, and error handling. Experiments across three benchmarks, models, and frameworks reveal that the choice of framework significantly impacts performance, rivaling the influence of the underlying LLM.
Framework choice in multi-agent systems matters just as much as the LLM itself, a fact obscured by existing model-centric benchmarks.
The rapid adoption of LLM-based agentic systems has produced a rich ecosystem of frameworks (smolagents, LangGraph, AutoGen, CAMEL, LlamaIndex, i.a.). Yet existing benchmarks are model-centric: they fix the agentic setup and do not compare other system components. We argue that implementation decisions substantially impact performance, including choices such as topology, orchestration logic, and error handling. MASEval addresses this evaluation gap with a framework-agnostic library that treats the entire system as the unit of analysis. Through a systematic system-level comparison across 3 benchmarks, 3 models, and 3 frameworks, we find that framework choice matters as much as model choice. MASEval allows researchers to explore all components of agentic systems, opening new avenues for principled system design, and practitioners to identify the best implementation for their use case. MASEval is available under the MIT licence https://github.com/parameterlab/MASEval.