parallel-web-search（并行网络搜索）

Parallel Web Search

Fast web search via Parallel AI REST API. Use for quick facts, news, finding sources. Results in 5-60 seconds.

CLI alternative (recommended): Install parallel-cli for official skill: npx skills add parallel-web/parallel-agent-skills --skill parallel-web-search

Setup

# Load API key (Windows/Git Bash compatible)
API_KEY=$(grep "^PARALLEL_API_KEY=" .env.local | cut -d= -f2)

Get your key at https://platform.parallel.ai — add to .env.local:

PARALLEL_API_KEY=your-key-here

Search Request

curl -s -X POST "https://api.parallel.ai/v1beta/search" \
  -H "x-api-key: $API_KEY" \
  -H "Content-Type: application/json" \
  -H "parallel-beta: search-extract-2025-10-10" \
  -d '{
    "objective": "your search query here"
  }'

Full Options

{
  "objective": "Find current Bitcoin price",
  "search_queries": ["BTC price today"],
  "max_results": 5,
  "mode": "agentic",
  "source_policy": {
    "include_domains": ["bloomberg.com", "reuters.com"]
  }
}

Response

{
  "results": [
    {
      "title": "...",
      "url": "...",
      "excerpt": "...",
      "relevance_score": 0.95
    }
  ]
}

Source Filtering

Include only specific domains:

{"source_policy": {"include_domains": ["arxiv.org", "nature.com"]}}

Exclude domains:

{"source_policy": {"exclude_domains": ["reddit.com", "twitter.com"]}}

When to Use

• Quick fact lookups, current prices, news
• Finding source URLs for further extraction
• Research queries where speed > depth

For deep analysis, use parallel-deep-research instead.

Error Handling