MiniWob++ (Mini World of Bits++) is a collection of 125 browser-based web-interaction tasks ranging from simple button clicks to multi-step form filling. The CUBE wrapper starts a local HTTP server that serves the MiniWoB task pages and exposes them via a browser tool; each task runs independently with no shared state.
By: @younik (Omar Younis) , @amanjaiswal73892 (Aman Jaiswal) , @NicolasAG (Nicolas Gontier)
Install
pip install miniwob-cube
Version: 1.0.0 · PyPI page
Feature Flags
Legal
Reproducibility journal
How to submit →This is a reproducibility journal — not a leaderboard.
Submissions document how reference agents and models score over time, across infrastructures, cube versions, and package versions. Use it to detect drift and validate environments. Not a place to publish a new agent or fine-tune to "win" — there is no ranking, scores are self-reported, and submissions are unverified. To showcase a new agent or model, use ATLAS / EEE / your own benchmark page.
No submissions yet. Be the first — see how to submit.
Parallelization
task-parallel
Registry Entry (YAML)
id: miniwob
name: "MiniWob++"
version: "1.0.0"
description: >
MiniWob++ (Mini World of Bits++) is a collection of 125 browser-based
web-interaction tasks ranging from simple button clicks to multi-step form
filling. The CUBE wrapper starts a local HTTP server that serves the MiniWoB
task pages and exposes them via a browser tool; each task runs independently
with no shared state.
package: miniwob-cube
dev_install_url: "git+https://github.com/The-AI-Alliance/cube-harness#subdirectory=cubes/miniwob"
authors:
- github: younik
name: Omar Younis
- github: amanjaiswal73892
name: Aman Jaiswal
- github: NicolasAG
name: Nicolas Gontier
legal:
wrapper_license: MIT
benchmark_license:
reported: MIT
source_url: "https://github.com/Farama-Foundation/miniwob-plusplus/blob/master/LICENSE"
verified_by_original_authors: false
tags:
- web
- gui
paper: "https://arxiv.org/abs/2306.06070"
getting_started_url: "https://miniwob.farama.org"
parallelization_mode: task-parallel
# Fields below are set by CI — do not edit manually.
status: degraded
task_count: 125
has_debug_task: true
has_debug_agent: true
resources: []
features:
async: false
streaming: false
multi_agent: false
multi_dim_reward: false
action_space: []