Towards Resource-Efficient Multimodal Intelligence: Learned Routing among Specialized Expert Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Saini, Mayank, Bishwas, Arit Kumar
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911256180424704
author Saini, Mayank
Bishwas, Arit Kumar
author_facet Saini, Mayank
Bishwas, Arit Kumar
contents As AI moves beyond text, large language models (LLMs) increasingly power vision, audio, and document understanding; however, their high inference costs hinder real-time, scalable deployment. Conversely, smaller open-source models offer cost advantages but struggle with complex or multimodal queries. We introduce a unified, modular framework that intelligently routes each query - textual, multimodal, or complex - to the most fitting expert model, using a learned routing network that balances cost and quality. For vision tasks, we employ a two-stage open-source pipeline optimized for efficiency and reviving efficient classical vision components where they remain SOTA for sub-tasks. On benchmarks such as Massive Multitask Language Understanding (MMLU) and Visual Question Answering (VQA), we match or exceed the performance of always-premium LLM (monolithic systems with one model serving all query types) performance, yet reduce the reliance on costly models by over 67%. With its extensible, multi-agent orchestration, we deliver high-quality, resource-efficient AI at scale.
format Preprint
id arxiv_https___arxiv_org_abs_2511_06441
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Towards Resource-Efficient Multimodal Intelligence: Learned Routing among Specialized Expert Models
Saini, Mayank
Bishwas, Arit Kumar
Computation and Language
Machine Learning
I.2.7; I.2.6; I.2.11
As AI moves beyond text, large language models (LLMs) increasingly power vision, audio, and document understanding; however, their high inference costs hinder real-time, scalable deployment. Conversely, smaller open-source models offer cost advantages but struggle with complex or multimodal queries. We introduce a unified, modular framework that intelligently routes each query - textual, multimodal, or complex - to the most fitting expert model, using a learned routing network that balances cost and quality. For vision tasks, we employ a two-stage open-source pipeline optimized for efficiency and reviving efficient classical vision components where they remain SOTA for sub-tasks. On benchmarks such as Massive Multitask Language Understanding (MMLU) and Visual Question Answering (VQA), we match or exceed the performance of always-premium LLM (monolithic systems with one model serving all query types) performance, yet reduce the reliance on costly models by over 67%. With its extensible, multi-agent orchestration, we deliver high-quality, resource-efficient AI at scale.
title Towards Resource-Efficient Multimodal Intelligence: Learned Routing among Specialized Expert Models
topic Computation and Language
Machine Learning
I.2.7; I.2.6; I.2.11
url https://arxiv.org/abs/2511.06441