Install
openclaw skills install nano-web-searchAI-powered web search via NanoGPT API with multiple providers (linkup, tavily, exa, kagi, perplexity, valyu, brave). Supports searchResults, sourcedAnswer, a...
openclaw skills install nano-web-searchAI-powered web search via NanoGPT API with multiple providers.
python3 - For payload building and output formattingcurl - For HTTP requestsrequests Python package - For Python API usageInstall the requests package:
pip install requests
Set your NanoGPT API key as an environment variable:
export NANOGPT_API_KEY="sk-nano-your-key-here"
Get your API key from nano-gpt.com.
# Basic search
./search.sh "your query"
# With specific provider
./search.sh "AI trends 2025" --provider tavily
# Deep search
./search.sh "quantum computing" --depth deep
# Search with date filter
./search.sh "climate news" --from-date 2025-01-01 --to-date 2025-02-18
# Sourced answer (synthesized response with citations)
./search.sh "what is RAG" --output-type sourcedAnswer
# JSON output (for scripting)
./search.sh "python tutorials" --json
linkup (default) - $0.006 standard, $0.06 deeptavily - $0.008 standard, $0.016 deepexa - $0.005 base + per-pagekagi - $0.002 web/news, $0.025 searchperplexity - $0.005 flatvalyu - ~$0.0015/resultbrave - $0.005 flat--provider - Search provider (default: linkup)--depth - Search depth: standard or deep--output-type - Output format: searchResults, sourcedAnswer, structured--from-date - Start date filter (YYYY-MM-DD)--to-date - End date filter (YYYY-MM-DD)--include-domains - Comma-separated domains to include--exclude-domains - Comma-separated domains to exclude--max-results - Max results (provider-specific)--json - Output raw JSONfrom search import NanoWebSearch
# Pass API key explicitly (recommended)
search = NanoWebSearch(api_key="sk-nano-your-key")
# Or use NANOGPT_API_KEY environment variable
search = NanoWebSearch()
results = search.search("AI trends", provider="tavily", depth="deep")
for r in results['data']:
print(f"{r['title']}: {r['url']}")