Cutscene Agent: An LLM Agent Framework for Automated 3D Cutscene Generation

2026-04-28

AI Summary

Traditional 3D cinematic production relies on labor-intensive collaboration among multidisciplinary experts, resulting in cumbersome workflows and high costs. This work proposes a large language model–based multi-agent framework that enables end-to-end generation of editable, engine-native 3D cinematics directly from screenplays. A director agent orchestrates specialized sub-agents responsible for animation, camera, and sound design, iteratively refining outputs through visual feedback. Key contributions include Cutscene Toolkit—the first game engine toolkit supporting bidirectional integration via the Model Context Protocol—a novel multi-agent architecture with visual reasoning capabilities, and CutsceneBench, the first benchmark for evaluating long-horizon, multi-step tool-use tasks. Experiments demonstrate the framework’s ability to generate high-quality cinematics and provide systematic evaluation of mainstream large language models on complex, extended tasks.

Abstract

Cutscenes are carefully choreographed cinematic sequences embedded in video games and interactive media, serving as the primary vehicle for narrative delivery, character development, and emotional engagement. Producing cutscenes is inherently complex: it demands seamless coordination across screenwriting, cinematography, character animation, voice acting, and technical direction, often requiring days to weeks of collaborative effort from multidisciplinary teams to produce minutes of polished content. In this work, we present Cutscene Agent, an LLM agent framework for automated end-to-end cutscene generation. The framework makes three contributions: (1)~a Cutscene Toolkit built on the Model Context Protocol (MCP) that establishes \emph{bidirectional} integration between LLM agents and the game engine -- agents not only invoke engine operations but continuously observe real-time scene state, enabling closed-loop generation of editable engine-native cinematic assets; (2)~a multi-agent system where a director agent orchestrates specialist subagents for animation, cinematography, and sound design, augmented by a visual reasoning feedback loop for perception-driven refinement; and (3)~CutsceneBench, a hierarchical evaluation benchmark for cutscene generation. Unlike typical tool-use benchmarks that evaluate short, isolated function calls, cutscene generation requires long-horizon, multi-step orchestration of dozens of interdependent tool invocations with strict ordering constraints -- a capability dimension that existing benchmarks do not cover. We evaluate a range of LLMs on CutsceneBench and analyze their performance across this challenging task.

Problem

• cutscene generation
• automated animation
• LLM agent
• multi-agent coordination
• long-horizon task

Innovation

• LLM agent
• cutscene generation
• Model Context Protocol
• multi-agent system
• visual reasoning