mlsanityData to 14 Oct 2026

INAR-VL: Input-Aware Routing for Edge-Cloud Vision-Language Inference

arXiv 2605.18853 ↗First posted 13 May 2026cs.LG cs.CV cs.DCcited by 2

On the map: Multimodal LLMs / vision-language models → Visual token pruning and efficiency

Abstract (arXiv)

Edge deployment of Vision-Language Models (VLMs) faces a tradeoff between latency and accuracy: cloud execution provides high-quality predictions but incurs communication delay and energy cost, while edge-only execution is faster but less accurate due to limited model capacity. This trade-off is further complicated by heterogeneity in image quality and reasoning complexity, making static placement suboptimal. We present INAR-VL, a lightweight edge-cloud routing system for multimodal inference in a two-tier deployment. INAR-VL maintains complementary VLMs across edge and cloud and uses lightweight image and text complexity signals to guide routing and model selection, executing simple queries locally while offloading complex ones when beneficial. Evaluation on visual question answering shows that INAR-VL executes 36% of requests on the edge, reduces latency by 24%, lowers energy by 26%, and preserves 97% of cloud-level accuracy.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in Visual token pruning and efficiency, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 12 of its 17 references that are in the map. First posted on arXiv 13 May 2026.

Builds on

12 of 17 references.

Closest ancestors · 1

Same direction, newest first

  1. 2026Vision-Language Models on the Edge for Real-Time Robotic Perception21 Jan

Other references · 11

  1. 2016Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answeringcited by 2,516
  2. 2019Towards VQA Models That Can Readcited by 1,718
  3. 2023FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performancecited by 853
  4. 2024RouteLLM: Learning to Route LLMs with Preference Datacited by 682
  5. 2024Hybrid LLM: Cost-Efficient and Quality-Aware Query Routingcited by 372
  6. 2023Generating Images with Multimodal Language Modelscited by 260
  7. 2025LLaVA-OneVision-1.5: Fully Open Framework for Democratized Multimodal Trainingcited by 244
  8. 2024Efficient Multimodal Large Language Models: A Surveycited by 96
  9. 2024Smoothie: Label Free Language Model Routingcited by 31
  10. 2024CE-CoLLM: Efficient and Adaptive Large Language Models Through Cloud-Edge Collaborationcited by 25
  11. 2025Efficient Routing of Inference Requests across LLM Instances in Cloud-Edge Computingcited by 5

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026LAST: The Last Query Token Guides Visual Token Pruning for Edge-Cloud Collaborative MLLM Inferencecited by 0
  2. 2026MHR-Router: Multimodal Heterogeneous Retrieval-Enhanced Adaptive Routing for VLM Inferencecited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Progressive Semantic Communication for Efficient Edge-Cloud Vision-Language Models
  2. 2025Collaborative Edge-to-Server Inference for Vision-Language Models
  3. 2025Vision-Language Models for Edge Networks: A Comprehensive Survey
  4. 2026RAPID: Redundancy-Aware and Compatibility-Optimal Edge-Cloud Partitioned Inference for Diverse VLA Models
  5. 2025edgeVLM: Cloud-edge Collaborative Real-time VLM based on Context Transfer
  6. 2025ECVL-ROUTER: Scenario-Aware Routing for Vision-Language Models
  7. 2025A Survey on Efficient Vision-Language Models
  8. 2026Towards Efficient Large Vision-Language Models: A Comprehensive Survey on Inference Strategies