POST /report

Analyze an entire menu and get a comprehensive quality report: duplicates, noise detection, cuisine breakdown, non-food items, and an overall quality score. Useful for onboarding new restaurants or auditing existing catalog data.

Request

Parameter	Type	Required	Description
`items`	string[]	Yes	Full menu item list to analyze. 1-2000 items, max 500 chars each.
`restaurant_name`	string	No	Restaurant name for the report header. Default: empty string.

{
  "items": [
    "Chicken Biryani",
    "**NEW** Murgh Biryani Serves 2",
    "Paneer Tikka",
    "Panner Tika",
    "Masala Dosa",
    "Paper Napkins",
    "Cigarette Lighter"
  ],
  "restaurant_name": "Sample Restaurant"
}

Response

Field	Type	Description
`restaurant_name`	string	Restaurant name (if provided)
`total_items`	int	Total menu items analyzed
`noise_analysis`	object	Promotional text and formatting noise analysis
`dedup_summary`	object	Duplicate item detection summary
`cuisine_breakdown`	object or null	Cuisine distribution (null if classifier not loaded)
`non_food_items`	object[]	Items detected as non-food
`quality_score`	int	Overall menu quality score (0-100, higher is better)
`processing_time_ms`	float	Processing time in milliseconds

noise_analysis

Field	Type	Description
`items_with_noise`	int	Count of items with promotional or formatting noise
`noise_percentage`	float	Percentage of menu items with noise
`examples`	object[]	Up to 5 examples showing `original` and `cleaned` text

dedup_summary

Field	Type	Description
`total_clusters`	int	Number of duplicate groups found
`total_duplicate_items`	int	Total excess duplicates across all clusters
`top_clusters`	object[]	Largest duplicate clusters (up to 10)

cuisine_breakdown

Field	Type	Description
`distribution`	object	Cuisine label to item count mapping, e.g. `{"North Indian": 12, "Chinese": 5}`

non_food_items

Field	Type	Description
`item`	string	Item text flagged as non-food
`category`	string	Non-food category: pharmacy, grocery, or merchandise

Example

import requests

response = requests.post("https://dish-embed.latimal.com/report",
    headers={"X-API-Key": "YOUR_KEY", "Content-Type": "application/json"},
    json={
        "items": menu_items,
        "restaurant_name": "Biryani House"
    }
)

report = response.json()
print(f"Quality Score: {report['quality_score']}/100")
print(f"Duplicates: {report['dedup_summary']['total_duplicate_items']} excess items")
print(f"Noise: {report['noise_analysis']['noise_percentage']:.0f}% of items have noise")
print(f"Non-food: {len(report['non_food_items'])} items flagged")

Availability

Available on the Pro and Scale plans. Each request counts as one API call against your monthly quota, regardless of how many items you send.

Try it live — Test this endpoint in the interactive playground.

For a complete integration walkthrough, see the Menu Health Monitoring guide.

Notes

This endpoint runs the full analysis pipeline (embed + dedup + classify + noise detection + non-food detection), so it does more work per call and has higher latency than the individual endpoints.
The quality score penalizes: high duplicate percentage, noise, non-food items, and low cuisine diversity.
Non-food detection catches items like paper napkins, cigarettes, phone chargers, and other merchandise that sometimes appears in delivery platform catalogs.
For menus over 2000 items, split into batches.

On this page