Beyond Gemini-3-Pro: Revisiting LLM Routing and Aggregation at Scale

2026-01-04

arXiv.org

AI Summary

Current collaborative frameworks for large language models are constrained by training-free query-dependent routing, static aggregation strategies, and insufficient exploitation of the complementary relationship between routing and aggregation. This work proposes a novel paradigm of dynamic collaborative routing and aggregation, introducing three key innovations: query-response hybrid routing, support-set-based aggregator selection, and adaptive routing-aggregation switching. These mechanisms jointly orchestrate ten open-source large language models in a coordinated manner. Evaluated across nine benchmarks, the proposed approach outperforms Gemini-3-Pro by a significant margin while achieving a 47% cost advantage, and it consistently surpasses established baselines. The results underscore the promise of collective intelligence as a viable pathway toward artificial general intelligence.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Planning, Routing, and Scheduling: Planning with Language Models
• Natural Language Processing: (Large) Language Models

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Large Language Models (LLMs) have rapidly advanced, with Gemini-3-Pro setting a new performance milestone. In this work, we explore collective intelligence as an alternative to monolithic scaling, and demonstrate that open-source LLMs'collaboration can surpass Gemini-3-Pro. We first revisit LLM routing and aggregation at scale and identify three key bottlenecks: (1) current train-free routers are limited by a query-based paradigm focusing solely on textual similarity; (2) recent aggregation methods remain largely static, failing to select appropriate aggregators for different tasks;(3) the complementarity of routing and aggregation remains underutilized. To address these problems, we introduce JiSi, a novel framework designed to release the full potential of LLMs'collaboration through three innovations: (1) Query-Response Mixed Routing capturing both semantic information and problem difficulty; (2) Support-Set-based Aggregator Selection jointly evaluating the aggregation and domain capacity of aggregators; (3) Adaptive Routing-Aggregation Switch dynamically leveraging the advantages of routing and aggregation. Comprehensive experiments on nine benchmarks demonstrate that JiSi can surpass Gemini-3-Pro with only 47% costs by orchestrating ten open-source LLMs, while outperforming mainstream baselines. It suggests that collective intelligence represents a novel path towards Artificial General Intelligence (AGI).

Problem

• LLM routing
• aggregation
• collective intelligence
• query-response paradigm
• complementarity

Innovation

• LLM collaboration
• query-response mixed routing
• support-set-based aggregation
• adaptive routing-aggregation switch
• collective intelligence