# FlashRT: Agent Harness for Guiding Agents to Deploy Real-Time Multimodal Applications

arXiv: https://arxiv.org/abs/2607.18171 · HTML: https://mlsanity.com/paper/2607.18171/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Multimodal and embodied agents, one of 16 directions in LLM agents and tool use.
- Builds on 35 of its 54 references that are in the map.
- First posted on arXiv 20 Jul 2026.

## Placed in

- LLM agents and tool use / [Multimodal and embodied agents](https://mlsanity.com/agents/multimodal-and-embodied-agents/)

## Builds on (most cited first, 30 of 35)

- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [Reflexion: Language Agents with Verbal Reinforcement Learning](https://mlsanity.com/paper/2303.11366/) (2023)
- [π0: A Vision-Language-Action Flow Model for General Robot Control](https://mlsanity.com/paper/2410.24164/) (2024)
- [MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework](https://mlsanity.com/paper/2308.00352/) (2023)
- [SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering](https://mlsanity.com/paper/2405.15793/) (2024)
- [SGLang: Efficient Execution of Structured Language Model Programs](https://mlsanity.com/paper/2312.07104/) (2023)
- [OpenHands: An Open Platform for AI Software Developers as Generalist Agents](https://mlsanity.com/paper/2407.16741/) (2024)
- [AlphaEvolve: A coding agent for scientific and algorithmic discovery](https://mlsanity.com/paper/2506.13131/) (2025)
- [Qwen2.5-Omni Technical Report](https://mlsanity.com/paper/2503.20215/) (2025)
- [DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Serving](https://mlsanity.com/paper/2401.09670/) (2024)
- [Eureka: Human-Level Reward Design via Coding Large Language Models](https://mlsanity.com/paper/2310.12931/) (2023)
- [Moshi: a speech-text foundation model for real-time dialogue](https://mlsanity.com/paper/2410.00037/) (2024)
- [EMO: Emote Portrait Alive -- Generating Expressive Portrait Videos with Audio2Video Diffusion Model under Weak Conditions](https://mlsanity.com/paper/2402.17485/) (2024)
- [KernelBench: Can LLMs Write Efficient GPU Kernels?](https://mlsanity.com/paper/2502.10517/) (2025)
- [Hallo: Hierarchical Audio-Driven Visual Synthesis for Portrait Image Animation](https://mlsanity.com/paper/2406.08801/) (2024)
- [Beyond Data and Model Parallelism for Deep Neural Networks](https://mlsanity.com/paper/1807.05358/) (2018)
- [WorldPlay: Towards Long-Term Geometric Consistency for Real-Time Interactive World Modeling](https://mlsanity.com/paper/2512.14614/) (2025)
- [Advancing Open-source World Models](https://mlsanity.com/paper/2601.20540/) (2026)
- [GSPMD: General and Scalable Parallelization for ML Computation Graphs](https://mlsanity.com/paper/2105.04663/) (2021)
- [Learning Performance-Improving Code Edits](https://mlsanity.com/paper/2302.07867/) (2023)
- [SyncTalk: The Devil is in the Synchronization for Talking Head Synthesis](https://mlsanity.com/paper/2311.17590/) (2023)
- [Kevin: Multi-Turn RL for Generating CUDA Kernels](https://mlsanity.com/paper/2507.11948/) (2025)
- [CUDA-L1: Improving CUDA Optimization via Contrastive Reinforcement Learning](https://mlsanity.com/paper/2507.14111/) (2025)
- [StreamDiffusion: A Pipeline-level Solution for Real-time Interactive Generation](https://mlsanity.com/paper/2312.12491/) (2023)
- [Towards Robust Agentic CUDA Kernel Benchmarking, Verification, and Optimization](https://mlsanity.com/paper/2509.14279/) (2025)
- [Astra: A Multi-Agent System for GPU Kernel Performance Optimization](https://mlsanity.com/paper/2509.07506/) (2025)
- [CudaForge: An Agent Framework with Hardware Feedback for CUDA Kernel Optimization](https://mlsanity.com/paper/2511.01884/) (2025)
- [AutoTriton: Automatic Triton Programming with Reinforcement Learning in LLMs](https://mlsanity.com/paper/2507.05687/) (2025)
- [LongLive-2.0: An NVFP4 Parallel Infrastructure for Long Video Generation](https://mlsanity.com/paper/2605.18739/) (2026)

## Built on by (most cited first, 1 of 1)

- [Efficient Vision-Language-Action Management and Serving for Robot Factories](https://mlsanity.com/paper/2609.12075/) (2026)

## Abstract (arXiv)

Real-time multimodal applications, including voice agents and interactive video generation, compose heterogeneous models into pipelines whose efficient deployment requires application-specific decisions about placement, streaming, and intra-model parallelism. Existing serving systems and auto-parallelism compilers commit to limited transformations and fixed workload assumptions, so achieving high performance on a new application requires hand-crafting an efficient implementation. We present FlashRT, an agent harness that guides coding agents to lift simple developer-written reference implementations into optimized multi-GPU deployments that flexibly weigh target metrics like latency and throughput. Using a new chain-of-program paradigm, FlashRT directs a generic coding agent through a multi-pass transformation process where an agent transforms the reference into an intermediate representation (IR) to capture data dependencies and persistent-state scopes, validates this IR via a sequential interpreter, and performs static analyses to identify candidate transformations. Then, the agent iteratively implements, verifies, and benchmarks each candidate under a measurement-gated optimization loop to produce effective deployments that span different hardware budgets. Across various applications, including video world models and multimodal LLMs, FlashRT converts reference implementations into highly efficient deployments, delivering up to ~70x latency reduction and 2.8x throughput improvement on NVIDIA B200 GPUs. On AMD MI355X GPUs, FlashRT matches the peak latency reduction while increasing peak throughput improvement to 3.6x, demonstrating that agent-driven optimization can be more scalable on platforms with less mature expert optimization. In fact, for Qwen3-Omni text-to-audio inference, FlashRT reduces response latency by 65% compared to the expert vLLM-Omni implementation on AMD MI355X.