Exa Contents
Fetch text from known URLs. Default: compact verbosity (filters navigation/footer noise).
Token tip: The
/contentsendpoint only supports full text. For token-efficient highlights or summary, useexa-search --highlightsorexa-search --summaryinstead.
Setup
Set EXA_API_KEY environment variable. Get key from Exa Dashboard.
Verbosity Comparison
| Mode | Tokens | Description |
|------|--------|-------------|
| --compact (default) | least | Main content only, filters noise |
| --standard | more | More detail |
| --full | most | Complete content, all sections |
Note:
--verbosityand--exclude-sectionsonly take effect when Exa performs a live crawl. For cached pages (most common), these options are silently ignored and full text is returned regardless.
Usage
# Default: compact verbosity
./contents.sh "https://example.com"
./contents.sh "https://url1.com" "https://url2.com"
# Control verbosity
./contents.sh --standard "https://example.com"
./contents.sh --full "https://example.com"
# Exclude noisy sections (auto-enables livecrawl)
./contents.sh --exclude-sections "navigation,footer,sidebar" "https://example.com"
# Force live crawl
./contents.sh --livecrawl "https://example.com"
Options
| Flag | Description |
|------|-------------|
| --compact | Compact verbosity (default) |
| --standard | Standard verbosity |
| --full | Full verbosity |
| --verbosity LEVEL | compact/standard/full |
| --exclude-sections S | Comma-separated: navigation,footer,sidebar,banner,header,metadata |
| --livecrawl | Force live crawl instead of cache |
For more details, see Exa Docs.