Web Search
Web search and page extraction from Tavily, Serper and Brave with the same API key and credits - from LLM grounding context to Google results and full page text.
Which one to use
| You need | Model | Credits |
|---|---|---|
| Context for an LLM or agent (with an answer) | tavily/search | 19 |
| LLM context sized to a token budget | brave/llm-context | 12 |
| Google results as-is | serper/search, serper/news, serper/images, serper/videos, serper/places, serper/shopping, serper/scholar | 3–5 |
| Results pages from an independent index | brave/web-search | 12 |
| Full text of pages you already have | tavily/extract, serper/scrape | 4–5 |
Model comparison
| Model | Credits | Type | Billing unit |
|---|---|---|---|
| serper/images | 3 | Search results (SERP) | per search (num > 10 reserved at 5, settled on Serper's charge) |
| serper/news | 3 | Search results (SERP) | per search (num > 10 reserved at 5, settled on Serper's charge) |
| serper/places | 3 | Search results (SERP) | per search |
| serper/scholar | 3 | Search results (SERP) | per search |
| serper/search | 3 | Search results (SERP) | per search (num > 10 reserved at 5, settled on Serper's charge) |
| serper/videos | 3 | Search results (SERP) | per search (num > 10 reserved at 5, settled on Serper's charge) |
| tavily/extract | 4 | Page extraction | per extracted URL (3.72/URL basic, 7.44 advanced; failed URLs refunded) |
| serper/scrape | 5 | Page extraction | per page (2 Serper credits) |
| serper/shopping | 5 | Search results (SERP) | per search (2 Serper credits) |
| brave/llm-context | 12 | LLM grounding | per request |
| brave/web-search | 12 | Search results (SERP) | per request |
| tavily/search | 19 | LLM grounding | per search (basic/fast/ultra-fast; 38 for advanced) |
- Tavily and Serper reserve the maximum when the request starts and settle on the usage the provider reports in the response, refunding the difference. Brave is flat per request.
- Results are the provider's JSON as-is; image and page links in them are not copied to your storage.
- Called through the same predictions API as every model (POST /v1/predictions → GET /v1/predictions/{id}). These are synchronous upstream, so jobs usually complete in 1-3 seconds.
Model details
Tavily Extract
Tavily
Extract clean, LLM-ready content (markdown or text) from up to 20 web pages in one call, optionally reranked to a query. Billed per successfully extracted URL - about 4 credits per URL (basic), failed URLs are refunded.
Serper Google Images
Serper
Google Images results through Serper - image URL, thumbnail, size, source page and title. 3 credits per search.
Brave LLM Context
Brave
Web search for AI agents from Brave's independent index - returns pre-extracted, relevance-ranked snippets per source, sized to a token budget, ready to drop into an LLM prompt. 12 credits per request.
Serper Google News
Serper
Google News results through Serper - headline, source, date, snippet and image for each article. 3 credits per search.
Serper Google Places
Serper
Google Maps places through Serper - name, address, rating, review count, category, phone and coordinates. 3 credits per search.
Serper Google Scholar
Serper
Google Scholar results through Serper - paper title, authors and venue, snippet, year, citation count and PDF link. 3 credits per search.
Serper Scrape
Serper
Fetch one web page through Serper - clean text (optionally markdown), page metadata and JSON-LD. 5 credits per page.
Serper Google Search
Serper
Google web search results through Serper - organic results, knowledge graph, answer box, People Also Ask and related searches. 3 credits per search.
Tavily Search
Tavily
Real-time web search built for LLMs and agents - returns ranked results with relevant content snippets, an optional LLM-generated answer, raw page content, images and domain/date filters. 19 credits per basic search, 38 per advanced search.
Serper Google Shopping
Serper
Google Shopping results through Serper - product title, price, seller, rating and image. 5 credits per search.
Serper Google Videos
Serper
Google Videos results through Serper - title, link, channel, duration, date and thumbnail. 3 credits per search.
Brave Web Search
Brave
Search results pages from Brave's independent index - web results plus news, videos, discussions, FAQ and infobox clusters, with country, language, freshness and Goggles controls. 12 credits per request.
Quick start
Tavily Search - five results and a short answer (19 credits):
curl -X POST https://api.core.today/v1/predictions \
-H "Content-Type: application/json" \
-H "X-API-Key: cdt_your_api_key" \
-d '{
"model": "tavily/search",
"input": {
"query": "latest AI video generation models 2026",
"max_results": 5,
"include_answer": "basic"
}
}'Brave LLM Context - prompt-ready snippets within 4,096 tokens (12 credits):
curl -X POST https://api.core.today/v1/predictions \
-H "Content-Type: application/json" \
-H "X-API-Key: cdt_your_api_key" \
-d '{
"model": "brave/llm-context",
"input": {
"q": "latest AI video generation models",
"maximum_number_of_tokens": 4096
}
}'Serper - Google results for Korea (3 credits):
curl -X POST https://api.core.today/v1/predictions \
-H "Content-Type: application/json" \
-H "X-API-Key: cdt_your_api_key" \
-d '{
"model": "serper/search",
"input": { "q": "생성형 AI API", "gl": "kr", "hl": "ko" }
}'Fetch the result with the returned id (result is the provider's JSON):
curl https://api.core.today/v1/predictions/{id} \
-H "X-API-Key: cdt_your_api_key"