Docs API Reference SDK Reference Agentic Usage AI Skills

Web Search

Crawl webpages

Explore a website link graph from a root URL and extract content from successful pages. Authenticate with a Bearer API key, or pass api_key in the JSON body as a compatibility fallback; both https://api.bazaarlink.ai/v1 and https://bazaarlink.ai/api/v1 are supported. Unknown fields are ignored. The response includes successful pages, response_time, usage.credits, our additive usage.cost (credits multiplied by the admin-set per-credit price), and our request_id. Tavily accepts timeout values from 10 to 150 seconds, but BazaarLink's effective maximum is 90 seconds; values above 90 are lowered to 90 so the call stays below the 100-second edge limit. Crawl credits are mapping cost plus extraction cost: 10 basic pages cost 1 + 2 = 3 credits; advanced extraction doubles the extraction part, and instructions double the mapping part. Failed pages are free. Actual credits draw from the shared daily free allowance; if the estimate from limit exceeds the remaining allowance, the excess is billed from balance, or the call returns 432 without balance and suggests lowering limit. Credits are mapping cost plus extraction cost: 10 basic pages cost 1 + 2 = 3 credits; advanced extraction doubles the extraction part, and instructions double the mapping part. Failed pages are free. Actual credits draw from the shared daily free allowance; if the estimate from limit exceeds the remaining allowance, the excess is billed from balance, or the call returns 432 without balance and suggests lowering limit. SDK examples; keep the key in an environment variable or another secure server-side setting. Python: from tavily import TavilyClient client = TavilyClient(api_key="...", api_base_url="https://api.bazaarlink.ai/v1") response = client.crawl("docs.tavily.com", instructions="Find Python SDK pages") JavaScript: import { tavily } from "@tavily/core" const tvly = tavily({ apiKey: process.env.BAZAARLINK_API_KEY, apiBaseURL: "https://api.bazaarlink.ai/v1" }) const response = await tvly.crawl("docs.tavily.com", { instructions: "Find Python SDK pages" }) LangChain (crawl only): from langchain_tavily import TavilyCrawl tool = TavilyCrawl(api_key="...", api_base_url="https://api.bazaarlink.ai/v1") response = tool.invoke({"url": "docs.tavily.com", "instructions": "Find Python SDK pages"})

POST/v1/crawl

Authorizations

Authorizationrequired

string · header

API key as bearer token in the Authorization header.

Body

urlrequired

string

Root URL for the crawl; the scheme is optional, for example docs.tavily.com.

required; scheme optional

Example: "docs.tavily.com"

instructions

string

Natural-language crawl instructions. When present, chunks_per_source is available and the mapping part costs 2 credits per 10 successful pages.

Example: "Find Python SDK pages"

chunks_per_source

integer

Maximum short content chunks per source; available only with instructions, from 1 to 5, default 3.

1-5; only with instructions; default: 3

Example: 3

max_depth

integer

Maximum crawl depth, 1 to 5, default 1.

1-5; default: 1

Example: 1

max_breadth

integer

Maximum links followed per page at each level, 1 to 500, default 20.

1-500; default: 20

Example: 20

limit

integer

Total links processed before stopping; minimum 1, default 50.

minimum: 1; default: 50

Example: 50

select_paths

string[]

path selection regex patterns; at most 50 entries, with each pattern limited to 200 characters.

max 50 entries; max 200 characters per regex

Example: ["^/docs/.*"]

select_domains

string[]

domain selection regex patterns; at most 50 entries, with each pattern limited to 200 characters.

max 50 entries; max 200 characters per regex

Example: ["^docs\.example\.com$"]

exclude_paths

string[]

path exclusion regex patterns; at most 50 entries, with each pattern limited to 200 characters.

max 50 entries; max 200 characters per regex

Example: ["^/private/.*"]

exclude_domains

string[]

domain exclusion regex patterns; at most 50 entries, with each pattern limited to 200 characters.

max 50 entries; max 200 characters per regex

Example: ["^private\.example\.com$"]

allow_external

boolean

Whether to include external-domain links in the final results, default true.

default: true

Example: true

include_images

boolean

Whether to include images extracted from pages, default false.

default: false

Example: false

extract_depth

string

Extraction depth. advanced handles more tables and embedded content and doubles the extraction part of credits per 5 successful pages.

default: basic

basicadvanced

Example: "basic"

format

string

Format of extracted page content: markdown or plain text, default markdown.

default: markdown

markdowntext

Example: "markdown"

include_favicon

boolean

Whether to include a favicon URL for each result, default false.

default: false

Example: false

timeout

number

Seconds to wait for the crawl. Tavily's range is 10-150, but BazaarLink's effective maximum is 90 seconds; values above 90 are lowered to 90.

Tavily: 10-150; BazaarLink default/effective maximum: 90 seconds; values above 90 are lowered to 90

Example: 90

include_usage

boolean

Whether to include usage in the response, default false; when included it shows credits and usage.cost.

default: false

Example: false

POST /v1/crawl

1
2
3
4
5
6
7
curl https://api.bazaarlink.ai/v1/crawl   -H "Authorization: Bearer $BAZAARLINK_API_KEY"   -H "Content-Type: application/json"   -d '{
    "url": "docs.tavily.com",
    "instructions": "Find Python SDK pages",
    "limit": 20,
    "timeout": 90,
    "include_usage": true
  }'

Response examples

Crawl completed. results contains successful pages; usage.credits is the actual credit charge and usage.cost is the customer price using the admin-set per-credit rate.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
{
  "base_url": "docs.tavily.com",
  "results": [
    {
      "url": "https://docs.tavily.com/welcome",
      "raw_content": "Welcome to the docs",
      "favicon": "https://docs.tavily.com/favicon.ico"
    }
  ],
  "response_time": 1.23,
  "usage": {
    "credits": 3,
    "cost": 0.03
  },
  "request_id": "crawl_123e4567-e89b-12d3-a456-426614174111"
}