Papers
arxiv:2505.18134

VideoGameBench: Can Vision-Language Models complete popular video games?

Published on May 23
· Submitted by ofirpress on May 28
Authors:
,
,

Abstract

VideoGameBench evaluates vision-language models' abilities in real-time video game interaction using only visual inputs and high-level objectives, highlighting challenges in human-like skills.

AI-generated summary

Vision-language models (VLMs) have achieved strong results on coding and math benchmarks that are challenging for humans, yet their ability to perform tasks that come naturally to humans--such as perception, spatial navigation, and memory management--remains understudied. Real video games are crafted to be intuitive for humans to learn and master by leveraging innate inductive biases, making them an ideal testbed for evaluating such capabilities in VLMs. To this end, we introduce VideoGameBench, a benchmark consisting of 10 popular video games from the 1990s that VLMs directly interact with in real-time. VideoGameBench challenges models to complete entire games with access to only raw visual inputs and a high-level description of objectives and controls, a significant departure from existing setups that rely on game-specific scaffolding and auxiliary information. We keep three of the games secret to encourage solutions that generalize to unseen environments. Our experiments show that frontier vision-language models struggle to progress beyond the beginning of each game. We find inference latency to be a major limitation of frontier models in the real-time setting; therefore, we introduce VideoGameBench Lite, a setting where the game pauses while waiting for the LM's next action. The best performing model, Gemini 2.5 Pro, completes only 0.48% of VideoGameBench and 1.6% of VideoGameBench Lite. We hope that the formalization of the human skills mentioned above into this benchmark motivates progress in these research directions.

Community

Paper submitter
edited 6 days ago

We have a new benchmark that challenges frontier VLMs to play DOS and Game Boy games from the 1990s. The top LM (Gemini) completes just 0.48% of the games in the benchmark.

https://vgbench.com has lots of clips and info.

Paper author

We also have a code repository you can check out: https://github.com/alexzhang13/videogamebench

You can generate clips like these, e.g. Gemini 2.5 Pro playing Kirby's Dream Land:

This is an automated message from the Librarian Bot. I found the following papers similar to this paper.

The following papers were recommended by the Semantic Scholar API

Please give a thumbs up to this comment if you found it helpful!

If you want recommendations for any Paper on Hugging Face checkout this Space

You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: @librarian-bot recommend

Sign up or log in to comment

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2505.18134 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2505.18134 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2505.18134 in a Space README.md to link it from this page.

Collections including this paper 1