jeremylongshore SerpApi Core Workflow B: Multi-Engine Search
Overview
SerpApi supports 15+ search engines beyond Google. Each engine has its own parameters and result structure. Key engines: YouTube (search_query), Bing (q), Google News, Google Shopping, Google Maps, Walmart, eBay, Apple App Store.
Instructions
Step 1: YouTube Search
import serpapi, os
client = serpapi.Client(api_key=os.environ["SERPAPI_API_KEY"])
# YouTube uses search_query (not q)
yt = client.search(engine="youtube", search_query="python asyncio tutorial")
for video in yt.get("video_results", []):
print(f"{video['title']}")
print(f" Channel: {video.get('channel', {}).get('name')}")
print(f" Views: {video.get('views')}, Length: {video.get('length')}")
print(f" Link: {video['link']}")
print(f" Published: {video.get('published_date')}")
Step 2: Bing Search
bing = client.search(engine="bing", q="machine learning frameworks", count=10)
for r in bing.get("organic_results", []):
print(f"{r['position']}. {r['title']}")
print(f" {r['link']}")
# Bing has different snippet structure
print(f" {r.get('snippet', 'N/A')}")
Step 3: Google News
news = client.search(engine="google_news", q="artificial intelligence", gl="us", hl="en")
for article in news.get("news_results", []):
print(f"{article['title']}")
print(f" Source: {article['source']['name']}")
print(f" Date: {article.get('date')}")
print(f" Link: {article['link']}")
# News often has thumbnail
if "thumbnail" in article:
print(f" Image: {article['thumbnail']}")
Step 4: Google Shopping
shopping = client.search(
engine="google_shopping",
q="mechanical keyboard",
gl="us",
hl="en",
)
for product in shopping.get("shopping_results", []):
print(f"{product['title']}")
print(f" Price: {product.get('price')}")
print(f" Source: {product.get('source')}")
print(f" Rating: {product.get('rating')} ({product.get('reviews', 0)} reviews)")
print(f" Link: {product['link']}")
Step 5: Google Maps / Local
maps = client.search(
engine="google_maps",
q="pizza restaurants",
ll="@30.2672,-97.7431,14z", # Austin, TX coordinates + zoom
)
for place in maps.get("local_results", []):
print(f"{place['title']} - {place.get('rating', 'N/A')} stars ({place.get('reviews', 0)} reviews)")
print(f" Address: {place.get('address')}")
print(f" Phone: {place.get('phone')}")
print(f" Type: {place.get('type')}")
print(f" Hours: {place.get('operating_hours', {}).get('monday')}")
Step 6: Cross-Engine Comparison
def multi_search(query: str) -> dict:
"""Search across multiple engines for the same query."""
engines = [
{"engine": "google", "q": query},
{"engine": "bing", "q": query},
{"engine": "youtube", "search_query": query},
{"engine": "google_news", "q": query},
]
results = {}
for params in engines:
result = client.search(**params)
engine = params["engine"]
key = "organic_results" if engine != "youtube" else "video_results"
if engine == "google_news":
key = "news_results"
results[engine] = result.get(key, [])[:3]
return results # 4 API credits total
Error Handling
| Error | Engine | Solution |
|---|---|---|
search_query required |
YouTube | Use search_query not q |
No shopping_results |
Google Shopping | Query must be product-related |
Empty local_results |
Google Maps | Add ll parameter with coordinates |
count vs num |
Bing | Bing uses count, Google uses num |
Resources
Next Steps
For common errors, see serpapi-common-errors.
Source: jeremylongshore/claude-code-plugins-plus-skills → plugins/saas-packs/serpapi-pack/skills/serpapi-core-workflow-b/SKILL.md