Docs API Reference SDK Reference Agentic Usage AI Skills

Web Search

Extract web content

Extract raw web content from one URL or a batch of up to 20 URLs. Successful URLs appear in results and per-URL failures appear in failed_results; even when HTTP 200 returns an empty results array, check both arrays. Output order is not guaranteed. The endpoint supports query reranking, 1–5 chunks per source (only when query is provided), basic or advanced extraction, images, favicons, markdown or text format, and a 1–60 second timeout. Unknown request fields are ignored. Use Authorization: Bearer <key>, or pass api_key in the JSON body as a Tavily-compatibility fallback. Both BazaarLink base URLs are supported: https://api.bazaarlink.ai/v1 and https://bazaarlink.ai/api/v1. Every 5 successful URL extractions costs 1 credit with basic or 2 credits with advanced; failed URLs are free. The customer price is credits multiplied by the per-credit price set by the admin: see the BazaarLink pricing page and use usage.cost from the response. The free allowance shares the same daily credits as search; when it is exhausted with no balance, the response is 432. cURL: `curl https://api.bazaarlink.ai/v1/extract -H "Authorization: Bearer $BAZAARLINK_API_KEY" -H "Content-Type: application/json" -d '{"urls":["https://example.com/article"]}'` Python: `TavilyClient(api_key=..., api_base_url="https://api.bazaarlink.ai/v1").extract(urls=[...])` JavaScript: `tavily({ apiKey, apiBaseURL })` LangChain: `TavilyExtract(api_base_url=...)`

POST/v1/extract

Authorizations

Authorizationrequired

string · header

API key as bearer token in the Authorization header.

Body

urlsrequired

string | string[]

URLs to extract. Pass one string or an array containing 1–20 URLs. An empty array or no valid URL returns 400.

string or array of 1–20 URLs

Example: ["https://example.com/article"]

query

string

User intent used to rerank extracted content chunks.

Example: "key findings"

chunks_per_source

integer

Maximum relevant content chunks returned per source. Must be 1–5 and is available only when query is provided.

1–5; only with query

Example: 3

extract_depth

string

Extraction depth. advanced can retrieve tables and embedded content with higher latency; basic costs 1 credit per 5 successful URLs and advanced costs 2 credits.

default: basic

basicadvanced

Example: "basic"

include_images

boolean

Whether to include image URLs extracted from each successful page.

default: false

Example: true

include_favicon

boolean

Whether to include the favicon URL in each successful result.

default: false

Example: true

format

string

Format of extracted content. markdown preserves Markdown structure; text returns plain text.

default: markdown

markdowntext

Example: "markdown"

timeout

number

Maximum seconds to wait for extraction, from 1–60. When omitted, the default is 10 seconds for basic and 30 seconds for advanced.

1–60 seconds; default depends on extract_depth

Example: 30

include_usage

boolean

Whether to include credits and usage.cost in the response.

default: false

Example: true

POST /v1/extract

1
2
3
4
5
6
7
8
9
10
11
12
13
14
curl https://api.bazaarlink.ai/v1/extract \
  -H "Authorization: Bearer $BAZAARLINK_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "urls": ["https://example.com/article", "https://example.org/docs"],
    "query": "key findings",
    "chunks_per_source": 3,
    "extract_depth": "basic",
    "include_images": true,
    "include_favicon": true,
    "format": "markdown",
    "timeout": 30,
    "include_usage": true
  }'

Response examples

Extraction completed. Check both results for successful items and failed_results for per-URL failures. HTTP 200 can have an empty results array, and output order is not guaranteed. usage.credits is credit usage; usage.cost is the USD customer price charged by BazaarLink, not upstream cost; request_id is the BazaarLink-generated request ID.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
{
  "results": [
    {
      "url": "https://example.com/article",
      "raw_content": "Article content extracted as Markdown.",
      "images": [
        "https://example.com/image.jpg"
      ],
      "favicon": "https://example.com/favicon.ico"
    }
  ],
  "failed_results": [
    {
      "url": "https://example.org/unavailable",
      "error": "Failed to retrieve content"
    }
  ],
  "response_time": 0.42,
  "usage": {
    "credits": 1,
    "cost": 0
  },
  "request_id": "bl-extract-example-request"
}