Parallel CLI

A Unix-friendly CLI for the Parallel Search and Extract APIs.

Semantic search and content extraction that understands intent, not just keywords — right from your terminal.

Install

bun install -g @rikalabs/parallel

Setup

Get your API key at parallel.ai, then:

parallel config set-key --key <your-key>

Search

parallel search --query "latest news about AI"

With agentic reasoning:

parallel search --query "compare React vs Vue" --mode agentic

Extract

parallel extract --url https://example.com

With a specific objective:

parallel extract --url https://example.com --objective "find pricing"

Batch Processing

# Multiple queries
parallel search --query "topic 1" --query "topic 2" --concurrency 5

# From stdin
cat urls.txt | parallel extract --stdin --concurrency 10

Options

--format json|text    Output format (default: json)
--pretty              Pretty print output
--max-results N       Number of results (default: 10)
--concurrency N       Parallel requests (default: 5)

Run parallel --help for full documentation.

Claude Code

Add web search to Claude Code:

mkdir -p ~/.claude/skills/parallel && curl -so ~/.claude/skills/parallel/SKILL.md https://raw.githubusercontent.com/Rika-Labs/parallel/main/SKILL.md

Contributing · MIT License