ROMA: Recursive Open Meta-Agent Framework for Long-Horizon Multi-Agent Systems

TL;DR

ROMA tackles long-horizon reasoning bottlenecks by introducing a recursive, domain-agnostic meta-agent with four roles: Atomizer, Planner, Executor, and Aggregator. It handles task decomposition and result aggregation in a dependency-aware MECE fashion, while controlling context growth through localized execution and aggregated artifacts. GEPA+ extends ROMA by jointly optimizing prompts across components, enabling task adaptation without fine-tuning and delivering consistent accuracy gains. Empirical results on SEAL-0, FRAMES, SimpleQA, and EQ-Bench show ROMA achieving leading system-level performance and matching or exceeding closed-source baselines in long-form generation tasks.

Abstract

Current agentic frameworks underperform on long-horizon tasks. As reasoning depth increases, sequential orchestration becomes brittle, context windows impose hard limits that degrade performance, and opaque execution traces make failures difficult to localize or debug. We introduce ROMA (Recursive Open Meta-Agents), a domain-agnostic framework that addresses these limitations through recursive task decomposition and structured aggregation. ROMA decomposes goals into dependency-aware subtask trees that can be executed in parallel, while aggregation compresses and validates intermediate results to control context growth. Our framework standardizes agent construction around four modular roles --Atomizer (which decides whether a task should be decomposed), Planner, Executor, and Aggregator -- which cleanly separate orchestration from model selection and enable transparent, hierarchical execution traces. This design supports heterogeneous multi-agent systems that mix models and tools according to cost, latency, and capability. To adapt ROMA to specific tasks without fine-tuning, we further introduce GEPA$+$, an improved Genetic-Pareto prompt proposer that searches over prompts within ROMA's component hierarchy while preserving interface contracts. We show that ROMA, combined with GEPA+, delivers leading system-level performance on reasoning and long-form generation benchmarks. On SEAL-0, which evaluates reasoning over conflicting web evidence, ROMA instantiated with GLM-4.6 improves accuracy by 9.9\% over Kimi-Researcher. On EQ-Bench, a long-form writing benchmark, ROMA enables DeepSeek-V3 to match the performance of leading closed-source models such as Claude Sonnet 4.5. Our results demonstrate that recursive, modular agent architectures can scale reasoning depth while remaining interpretable, flexible, and model-agnostic.

Figures (2)

• Figure 1: Overview of ROMA's recursive architecture. An Atomizer determines whether a task is atomic. Atomic tasks are executed directly, while non-atomic tasks are decomposed into subtasks by a Planner. Each subtask is executed recursively by Executors, after which an Aggregator merges the outputs of all descendants to produce the final result.
• Figure 2: ROMA's hierarchical execution flow. Non-atomic tasks are decomposed top-down through planning, with left-to-right dependencies guiding execution, while results are combined bottom-up through aggregation. Executors operate on atomic subtasks, producing intermediate outputs that are aggregated into higher-level artifacts; the inset shows a zoomed-in executor node with tool interaction.