Hugging Face's logo

Datasets:
AnyStackLabsdev
/
souslab-us-restaurant-menus

Tasks:
Text Classification
Table Question Answering
Text Retrieval
Modalities:
Tabular
Text
Formats:
csv
Languages:
English
Size:
10K - 100K
Tags:
food
restaurants
menus
commerce
dining
food-tech
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
1
souslab-us-restaurant-menus / README.md
anystackdev's picture
anystackdev
Add dataset card
95c59c3 verified
preview code
|
raw
history blame contribute delete
7.37 kB
metadata 
license: cc-by-nc-4.0
language:
  - en
task_categories:
  - text-classification
  - table-question-answering
  - text-retrieval
  - feature-extraction
  - text-generation
tags:
  - food
  - restaurants
  - menus
  - commerce
  - dining
  - food-tech
  - business
  - usa
  - location-data
  - structured-data
  - tabular
  - api-data
size_categories:
  - 10K<n<100K
pretty_name: Souslab — US Restaurant Menus (Sample)
configs:
  - config_name: default
    data_files:
      - split: train
        path: menu_items_sample.csv

Souslab — US Restaurant Menus

A structured sample of the Souslab US restaurant menu dataset: real restaurants, real menu items, real prices — normalized into a clean schema you can train on or analyze directly.

This sample is published openly under CC-BY-NC-4.0 for research and non-commercial evaluation. The full dataset — 449,000+ US restaurants and 44.3M+ menu items, refreshed continuously with chain-level aggregation — is available via the Souslab API under commercial license.

🚀 Quick start 

from datasets import load_dataset

ds = load_dataset("AnyStackLabsdev/souslab-us-restaurant-menus")
print(ds)
print(ds["train"][0])

Or download directly via Files and versions.

📊 What's in this sample

  • Restaurants: ~4,600 US restaurants
  • Menu items: ~10,000 menu items
  • Geographic coverage: All 50 US states + DC
  • Chain coverage: Independents and chain locations, with chain-level grouping
  • Snapshot date: 2026-05-29
  • Format: CSV
  • File size: ~2.1 MB

This is a representative sample — geographically and categorically balanced, but a fraction of the full Souslab corpus. The full dataset is ~4,400x larger.

🗂️ Schema

menu_items_sample.csv

Field Type Description
item_uid string Stable unique identifier for this menu item
restaurant_id string Foreign key — links to the restaurant
menu_name string Name of the menu this item appears on
section_name string Menu section or category (e.g. "Appetizers", "Drinks")
item_name string Display name of the menu item
description string | null Item description as listed on the menu
price float | null Price in local currency
price_text string | null Raw price string as scraped (e.g. "$12.99")
currency string ISO 4217 currency code (e.g. "USD")
out_of_stock boolean Whether the item was marked unavailable at time of snapshot
city_name string City of the restaurant
state_code string US state (2-letter code)
zip string Postal code
geo_lat float Latitude of the restaurant
geo_lng float Longitude of the restaurant
canonical_item_id string | null Souslab canonical item identifier for cross-restaurant deduplication

🧠 Methodology summary

The full methodology lives at souslab.site/methodology — covering collection sources, refresh cadence, chain matching, quality controls, and known gaps. Quick version:

  • Sources: restaurants' own published menus, public aggregator surfaces, and partner data feeds. We don't bypass authentication, scrape behind logins, or ingest user reviews or photos.
  • Chain matching: multi-signal (registry-anchored, name similarity, location/ownership context, menu signature) — same chain across locations resolves to the same chain_id. "Joe's Pizza" in NYC and "Joe's Pizza" in Topeka stay distinct unless ownership/menu signals tie them.
  • Canonical items: canonical_item_id links the same item appearing across multiple sources or snapshots — useful for deduplication and price tracking over time.
  • Refresh cadence (in the live API): tiered. Top chains daily, mid-tier 2–4× per week, high-traffic independents weekly, long-tail independents monthly.
  • Known gaps: thin coverage of dine-in independents without digital menus, no nutritional data, no promotional pricing modeling, US-only, English-only.

This sample represents a point-in-time snapshot. The live API is refreshed continuously.

🎯 Use cases

This dataset is designed for:

  • AI/ML training: fine-tuning LLMs on food/menu language, training cuisine and dietary classifiers, building food-aware vision/multimodal models, grounding RAG over real-world food knowledge
  • Market research: pricing analysis, menu trend tracking, regional cuisine prevalence, chain operational analysis
  • Product development: recommendation engines, menu search and discovery, dietary filtering, food delivery platforms
  • Academic research: food systems analysis, restaurant economics, regional dining patterns
  • Alt-data signal generation: restaurant inflation, regional pricing dynamics, chain operational shifts

For commercial deployment, real-time data, larger volumes, or the full 44M+ menu item universe, contact us for an API license.

⚖️ License

This sample is released under CC-BY-NC-4.0 — free for research and non-commercial use with attribution.

Commercial use of this sample, or of any derivative dataset or model trained on it, requires a commercial license from Any Stack Labs. This includes:

  • Production deployment in a commercial product or service
  • Training models for commercial use (including model-as-a-service)
  • Reselling, sublicensing, or distributing derivatives
  • Embedding in commercial datasets or research products

For commercial licensing, contact hello@souslab.site or visit souslab.site.

Attribution

When using this sample in research or other CC-BY-NC contexts, please attribute as:

"Data sourced from the Souslab US Restaurant Menus dataset by Any Stack Labs (souslab.site)."

📚 Citation 

@dataset{souslab_us_restaurant_menus_2026,
  title     = {Souslab US Restaurant Menus (Sample)},
  author    = {{Any Stack Labs}},
  year      = {2026},
  publisher = {Hugging Face},
  url       = {https://huggingface.co/datasets/AnyStackLabsdev/souslab-us-restaurant-menus},
  note      = {Sample of the full Souslab dataset; full dataset available via API at souslab.site}
}

🔗 Related

📬 Contact

📝 Changelog

  • v1.0 — 2026-05-29 — Initial public release. ~4,600 restaurants, ~10,000 menu items.

Souslab is built by Any Stack Labs.