# Evaluating Multimodal LLMs as Generalist Vision-Language-Action Agents for Drone Control: Commanding, Approaching, Tracking and Searching

arXiv: https://arxiv.org/abs/2609.01404 · HTML: https://mlsanity.com/paper/2609.01404/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Embodied and 3D vision-language models, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 29 of its 35 references that are in the map.
- First posted on arXiv 1 Sep 2026.

## Placed in

- Multimodal LLMs / vision-language models / [Embodied and 3D vision-language models](https://mlsanity.com/mllm/embodied-and-3d-vision-language-models/)
- Vision-language-action models / [Aerial and drone VLA](https://mlsanity.com/vla/aerial-and-drone-vla/)
- LLM agents and tool use / [Multimodal and embodied agents](https://mlsanity.com/agents/multimodal-and-embodied-agents/)

## Builds on (most cited first, 29 of 29)

- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [Improved Baselines with Visual Instruction Tuning](https://mlsanity.com/paper/2310.03744/) (2023)
- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [Qwen3-VL Technical Report](https://mlsanity.com/paper/2511.21631/) (2025)
- [Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond](https://mlsanity.com/paper/2308.12966/) (2023)
- [LLaVA-Video: Video Instruction Tuning With Synthetic Data](https://mlsanity.com/paper/2410.02713/) (2024)
- [AirSim: High-Fidelity Visual and Physical Simulation for Autonomous Vehicles](https://mlsanity.com/paper/1705.05065/) (2017)
- [RoboPoint: A Vision-Language Model for Spatial Affordance Prediction for Robotics](https://mlsanity.com/paper/2406.10721/) (2024)
- [PIVOT: Iterative Visual Prompting Elicits Actionable Knowledge for VLMs](https://mlsanity.com/paper/2402.07872/) (2024)
- [Cosmos 3: Omnimodal World Models for Physical AI](https://mlsanity.com/paper/2606.02800/) (2026)
- [MOKA: Open-World Robotic Manipulation through Mark-Based Visual Prompting](https://mlsanity.com/paper/2403.03174/) (2024)
- [AerialVLN: Vision-and-Language Navigation for UAVs](https://mlsanity.com/paper/2308.06735/) (2023)
- [UrbanVideo-Bench: Benchmarking Vision-Language Models on Embodied Intelligence with Video Data in Urban Spaces](https://mlsanity.com/paper/2503.06157/) (2025)
- [UAV-ON: A Benchmark for Open-World Object Goal Navigation with Aerial Agents](https://mlsanity.com/paper/2508.00288/) (2025)
- [CognitiveDrone: A VLA Model and Evaluation Benchmark for Real-Time Cognitive Task Solving and Reasoning in UAVs](https://mlsanity.com/paper/2503.01378/) (2025)
- [AirCopBench: A Benchmark for Multi-drone Collaborative Embodied Perception and Reasoning](https://mlsanity.com/paper/2511.11025/) (2025)
- [See, Point, Fly: A Learning-Free VLM Framework for Universal Unmanned Aerial Navigation](https://mlsanity.com/paper/2509.22653/) (2025)
- [BEDI: A Comprehensive Benchmark for Evaluating Embodied Agents on UAVs](https://mlsanity.com/paper/2505.18229/) (2025)
- [On Efficient Language and Vision Assistants for Visually-Situated Natural Language Understanding: What Matters in Reading and Reasoning](https://mlsanity.com/paper/2406.11823/) (2024)
- [RALLY: Role-Adaptive LLM-Driven Yoked Navigation for Agentic UAV Swarms](https://mlsanity.com/paper/2507.01378/) (2025)
- [Vision-and-Language Navigation for UAVs: Progress, Challenges, and a Research Roadmap](https://mlsanity.com/paper/2604.13654/) (2026)
- [Taking Flight with Dialogue: Enabling Natural Language Control for PX4-based Drone Agent](https://mlsanity.com/paper/2506.07509/) (2025)
- [AirHunt: Bridging VLM Semantics and Continuous Planning for Efficient Aerial Object Navigation](https://mlsanity.com/paper/2601.12742/) (2026)
- [Fly0: Persistent Metric Anchoring for Zero-Shot Aerial Vision-Language Navigation](https://mlsanity.com/paper/2602.15875/) (2026)
- [UAV-Track VLA: Embodied Aerial Tracking via Vision-Language-Action Models](https://mlsanity.com/paper/2604.02241/) (2026)
- [ESARBench: A Benchmark for Agentic UAV Embodied Search and Rescue](https://mlsanity.com/paper/2605.01371/) (2026)
- [State-Space Hierarchical Compression with Gated Attention and Learnable Sampling for Hour-Long Video Understanding in Large Multimodal Models](https://mlsanity.com/paper/2506.13564/) (2025)
- [DeTrack: A Benchmark and Altitude-Aware Dual World Model for Drone-embodied Tracking](https://mlsanity.com/paper/2605.17451/) (2026)
- [ActiveFly-Bench: Aligning Embodied Question Answering with Vision-Language-Action for Aerial Embodied Perception](https://mlsanity.com/paper/2607.10180/) (2026)

## Abstract (arXiv)

Multimodal large language models (MLLMs) are increasingly used to make decisions for drones, but existing evaluations are mostly limited to simulation, to a single drone, and to targets that are visible from the start and do not move. We present DroneCATS, a benchmark in which the MLLM alone controls the drone through four actions, and evaluate ten MLLMs from frontier models to small open models. The benchmark varies target visibility, target motion, scene and simulator, and extends to real drones. Beyond a single drone, it adds a multi-drone suite, where one model drives multiple drones, and a commander + sub-agents suite, where a commander model directs one sub-agent for each of the drones. Across these settings, we find that frontier models achieve the highest success rates, but their latency causes them to lose moving targets, while small open models often do not declare arrival even when they reach the target. Guided by this failure analysis, we fine-tune a 2B model on drone action data synthesized from public RGB-D images; its success rate rises from 0% to 31% in simulation and outperforms the frontier models in real flight. With multiple drones, we find that every zero-shot model succeeds more often as the commander of one sub-agent per drone than when it drives all of them from one context, and that commanding success follows general VLM ability. On real drones, we observe that the results broadly follow those in simulation, while latency and hardware limits matter more. We will fully open-source the benchmark.