SpeedrunBench: Challenging LLM Agents with Video Game Speedrunning

cs.AI

Submitted: 2026-10-06

Updated: 2026-10-06

Code: https://github.com/patronus-ai/speedrun_bench

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Sources

• ARC-AGI-3: A New Challenge for Frontier Agentic Intelligence
• Dota 2 with Large Scale Deep Reinforcement Learning
• Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities
• TextArena
• Mastering Diverse Domains through World Models
• PokeLLMon: A Human-Parity Agent for Pokemon Battles with Large Language Models
• VibeServe: Can AI Agents Build Bespoke LLM Serving Systems?
• The PokeAgent Challenge: Competitive and Long-Context Learning at Scale
• Continual Harness: Online Adaptation for Self-Improving Foundation Agents
• MuZero with Self-competition for Rate Control in VP9 Video Compression
• GameWorld: Towards Standardized and Verifiable Evaluation of Multimodal Game Agents
• Kimi K3: Open Frontier Intelligence
• Is Deep Reinforcement Learning Really Superhuman on Atari? Leveling the playing field
• StarCraft II: A New Challenge for Reinforcement Learning
• VideoGameBench: Can Vision-Language Models complete popular video games?

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• Agents' Last Exam
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems