WebShop
A simulated shopping environment with products and human instructions for training and evaluating web agents.
What the resource contains
- Product and instruction data used to populate the environment
- Human demonstration trajectories
- Search, navigation and simulated purchase actions
Potential uses
- Evaluate shopping web agents
- Research instruction following
- Compare imitation and reinforcement learning
Access & formats
JSON / JSONL / environment assets. External dependencies. Follow the project’s instructions on GitHub; some resources require external files, registration or approval.
License & reuse
LICENSE.md grants MIT rights to software and documentation. Underlying product content and other assets need separate consideration.
Important limitations
Purchases occur in a simulation. This benchmark is neither a payment integration nor a live merchant transaction feed.
Source & attribution
princeton-nlp — original GitHub project ↗
Documentation reviewed 2026-09-24. Review scope and licensing notes. Suggest a correction.
Related resources
Shopping MMLU
KL4805
A multitask benchmark for language models across shopping knowledge, reasoning, behavior and languages.
CSV / JSON · Repository / linked release
EcomInstruct Evaluation
Alibaba-NLP
Released evaluation tasks from EcomGPT covering product language understanding and generation.
JSON · Repository files
Amazon Product Text Evaluation
amazon-science
Product-description evaluation data for assessing language-model outputs against product features.
Text / structured metadata · Repository files