AI Paper Atlas

Research map / Multi-agent systems and games / LLM agents

MolmoWeb: Open Visual Web Agent and Open Data for the Open Web

Tanmay Gupta, Piper Wolters, Zixian Ma, Peter Sushko, Rock Yuren Pang, Yue Yang, Jason Ren, Harsh Trivedi, Taira Anderson, Winson Han, Ranjay Krishna

ECCV 2026 · Multi-agent systems and games › LLM agents · approach: Agentic benchmarks and orchestration · 25 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
scale training data from 1% to 10% (ablation)MolmoWeb · WebVoyager (30 steps)success rate (%): 44.5 -> 63.2Improved
scale training data from 10% to 100% (ablation)MolmoWeb · WebVoyager (30 steps)success rate (%): 63.2 -> 68.5Improved
add human trajectories to synthetic trajectories (ablation)MolmoWeb · WebVoyager (30 steps)success rate (%): 67.8 -> 68.5No clear effect
add human trajectories to synthetic trajectories (ablation)MolmoWeb · Online-Mind2Web (30 steps)success rate (%): 22.0 -> 21.4No clear effect
use synthetic AxTree-agent trajectories instead of human ones (same tasks, 2.7K) (ablation)MolmoWeb · WebVoyagersuccess rate (%): 35.4 -> 53.0Improved
use synthetic instead of human trajectories (2.7K) (ablation)MolmoWeb · Online-Mind2Websuccess rate (%): 9.0 -> 16.8Improved
use synthetic instead of human trajectories (2.7K) (ablation)MolmoWeb · DeepShopsuccess rate (%): 19.8 -> 24.4Improved
sample with top-p instead of greedy decoding (ablation)MolmoWeb · WebVoyagersuccess rate (%): 61.4 -> 68.5Improved
sample with top-p (p=0.8) instead of top-k (k=20) (ablation)MolmoWeb · WebVoyagersuccess rate (%): 67.4 -> 68.5Improved
use planner-operator-verifier multi-agent harness instead of Gemini AxTree agent alone for trajectory generation (ablation)Gemini AxTree agent · WebVoyagersuccess rate (%): 74.4 -> 78.5Improved
use parallel rollouts with best-of-4 selection instead of single rolloutMolmoWeb-8B · WebVoyagersuccess rate (%): 78.2 -> 94.7Improved
use parallel rollouts with best-of-4 selection instead of single rolloutMolmoWeb-8B · Online-Mind2Websuccess rate (%): 35.3 -> 60.5Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy