Deep search - Andi AI Search API Docs

Deep search explores your topic from multiple angles to find results that a single quick search might miss. It keeps searching until it finds strong, relevant matches — then corrects any typos in your query. Responses take approximately 2–3 seconds. Enable deep search by setting depth=deep:

curl "https://search-api.andisearch.com/api/v1/search?q=quantm+computing+applications&depth=deep" \
  -H "x-api-key: YOUR_API_KEY"

import requests

response = requests.get(
    "https://search-api.andisearch.com/api/v1/search",
    params={"q": "quantm computing applications", "depth": "deep"},
    headers={"x-api-key": "YOUR_API_KEY"}
)

data = response.json()

# Check if the query was spell-corrected
if data.get("correctedQuery"):
    print(f"Corrected: {data['correctedQuery']}")

for result in data["results"]:
    print(f"{result['title']} — {result['source']}")

const response = await fetch(
  "https://search-api.andisearch.com/api/v1/search?q=quantm+computing+applications&depth=deep",
  { headers: { "x-api-key": "YOUR_API_KEY" } }
);

const data = await response.json();

if (data.correctedQuery) {
  console.log(`Corrected: ${data.correctedQuery}`);
}

data.results.forEach(r => console.log(`${r.title} — ${r.source}`));

The example query intentionally misspells “quantum” — deep search corrects it automatically.

What deep search adds

Thorough coverage — finds results across multiple angles of your topic, covering sources a single search might miss
Higher quality results — keeps searching until it finds strong, relevant matches
Spell correction — fixes typos and misspellings in the query

When to use deep search

You need results a quick search might miss — niche topics, multi-faceted queries, or less obvious sources
Thoroughness matters more than speed
Queries come from user input that may contain typos
You want the highest quality results and can accept a 2–3 second response

Example response

{
  "results_type": "search",
  "answer": "",
  "type": "search",
  "title": "quantum computing applications",
  "results": [
    {
      "title": "Quantum Computing: Current Applications and Future Potential",
      "link": "https://example.com/quantum-applications",
      "desc": "Quantum computing is being applied in cryptography, drug discovery, and optimization problems...",
      "source": "example.com",
      "type": "website"
    }
  ],
  "correctedQuery": "quantum computing applications",
  "metrics": {
    "query": "quantum computing applications",
    "intent": "FallbackSearchIntent",
    "duration": 2340,
    "results_returned": 10,
    "total_results_found": 85
  }
}

Handling corrected queries

The correctedQuery field appears only when deep search detects and fixes a misspelling. Use it to show users what query was actually searched:

data = response.json()
if data.get("correctedQuery"):
    print(f"Showing results for: {data['correctedQuery']}")

correctedQuery is only present on deep search responses where a correction was made. Fast search does not perform spell correction.

Adding extracts and metadata

Combine deep search with richer output options:

# Deep search with text extracts
curl "https://search-api.andisearch.com/api/v1/search?q=quantum+computing&depth=deep&extracts=true" \
  -H "x-api-key: YOUR_API_KEY"

# Deep search with full metadata
curl "https://search-api.andisearch.com/api/v1/search?q=quantum+computing&depth=deep&metadata=full" \
  -H "x-api-key: YOUR_API_KEY"

# Deep search as LLM context
curl "https://search-api.andisearch.com/api/v1/search?q=quantum+computing&depth=deep&format=context" \
  -H "x-api-key: YOUR_API_KEY"

Next steps

Fast search

Default mode for real-time applications.

Response format

Response structure and result types.

Research assistant

Multi-query search with result aggregation.

Query parameters

Full parameter reference.

​What deep search adds

​When to use deep search

​Example response

​Handling corrected queries

​Next steps

Fast search

Response format

Research assistant

Query parameters

What deep search adds

When to use deep search

Example response

Handling corrected queries

Next steps