> For the complete documentation index see https://domscan.net/llms.txt. Markdown versions of every docs page are available by appending .md to the page URL.

Source: https://domscan.net/docs/similarity

# Domain Similarity

Explore Domain Similarity API documentation, request parameters, response fields, code examples, and error handling for DomScan integrations.

Compare two domains for visual and textual similarity using multiple algorithms including Levenshtein distance, Jaro-Winkler similarity, and visual homoglyph analysis. Useful for detecting typosquatting, brand impersonation, and phishing domains.

GET /v1/similarity

#### Query parameters

| Parameter | Type | Description |
| --- | --- | --- |
| domain1 required | string | First domain (typically the legitimate domain) |
| domain2 required | string | Second domain (potentially suspicious domain) |

#### Similarity algorithms

| Algorithm | Description |
| --- | --- |
| `levenshtein` | Edit distance normalized (0-1) |
| `jaro_winkler` | Prefix-weighted string similarity |
| `visual` | Homoglyph/lookalike character detection |

#### Example request

```bash
curl -H "X-API-Key: your-api-key" "https://domscan.net/v1/similarity?domain1=paypal.com&domain2=paypa1.com"
```

```python
import requests

domscan = requests.Session()
domscan.headers.update({"X-API-Key": "your-api-key"})

# Check multiple suspicious domains
legit = "paypal.com"
suspects = ["paypa1.com", "paypaI.com", "pаypal.com"]  # Note: last one has Cyrillic 'а'

for suspect in suspects:
    response = domscan.get(
        "https://domscan.net/v1/similarity",
        params={"domain1": legit, "domain2": suspect}
    )
    data = response.json()
    print(f"{suspect}: {data['typosquatting_risk']} risk (visual: {data['similarity']['visual']:.2f})")
```

#### Example response

```json
{
  "domain1": "paypal.com",
  "domain2": "paypa1.com",
  "similarity": {
    "levenshtein": 0.86,
    "jaro_winkler": 0.93,
    "visual": 0.95
  },
  "is_similar": true,
  "typosquatting_risk": "high",
  "homoglyphs_detected": ["l → 1"],
  "risk_factors": ["character_substitution", "high_visual_similarity"]
}
```

#### Response fields

| Field | Type |
| --- | --- |
| `domain1` | string |
| `domain2` | string |
| `overall_similarity` | number |
| `risk_level` | string |
| `analysis` | object |
| `analysis.levenshtein_distance` | integer |
| `analysis.levenshtein_similarity` | number |
| `analysis.jaro_winkler_similarity` | number |
| `analysis.visual_similarity` | number |
| `analysis.keyboard_distance` | number |
| `analysis.common_prefix_length` | integer |
| `analysis.common_suffix_length` | integer |
| `matching_methods[]` | string[] |
| `is_potential_typosquat` | boolean |
| `checked_at` | string |
| `meta` | object |