AndesVL Technical Report: An Efficient Mobile-side Multimodal Large Language Model

2025-10-13

AI Summary

To address the deployment challenges of mainstream multimodal large language models (MLLMs)—such as Qwen-VL and GPT-4o—on resource-constrained edge devices (e.g., smartphones) due to limited memory, power, and computational capacity, this paper introduces AndesVL, a lightweight, on-device MLLM architecture with 0.6B–4B parameters. Methodologically, AndesVL adopts an end-to-end training framework integrating the Qwen3 language model with multiple vision encoders, trained on large-scale multitask data and enhanced by an innovative 1+N LoRA fine-tuning strategy to enable efficient training and synergistic capability expansion. Evaluated across diverse benchmarks—including text-rich image understanding, mathematical reasoning, multi-image comprehension, visual question answering (VQA), hallucination mitigation, multilingual understanding, and GUI interpretation—AndesVL achieves state-of-the-art performance at comparable parameter counts, significantly advancing practical multimodal understanding capabilities on edge devices.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

In recent years, while cloud-based MLLMs such as QwenVL, InternVL, GPT-4o, Gemini, and Claude Sonnet have demonstrated outstanding performance with enormous model sizes reaching hundreds of billions of parameters, they significantly surpass the limitations in memory, power consumption, and computing capacity of edge devices such as mobile phones. This paper introduces AndesVL, a suite of mobile-side MLLMs with 0.6B to 4B parameters based on Qwen3's LLM and various visual encoders. We comprehensively outline the model architectures, training pipeline, and training data of AndesVL, which achieves first-tier performance across a wide range of open-source benchmarks, including fields such as text-rich image understanding, reasoning and math, multi-image comprehension, general VQA, hallucination mitigation, multilingual understanding, and GUI-related tasks when compared with state-of-the-art models of a similar scale. Furthermore, we introduce a 1+N LoR

Problem

• Developing efficient multimodal models for mobile devices
• Overcoming memory and computational constraints on edge devices
• Achieving competitive performance across diverse vision-language tasks

Innovation

• Mobile-side multimodal model with 0.6B-4B parameters
• Based on Qwen3 LLM and various visual encoders
• Achieves top performance across multiple benchmark categories