Web Research Parallel

Description

Runs web research via the Parallel API using OpenAI-compatible chat completions grounded in live web data. Supports a FindAll mode that converts a natural language objective into a structured async search spec, URL content extraction focused on a research objective, and task-based async runs for longer jobs.

What this skill does

1. 1Runs web searches using Parallel's neural search engine, returning OpenAI-compatible chat completions.
2. 2Supports a FindAll mode that transforms a natural language objective into a structured search spec and runs it asynchronously.
3. 3Extracts relevant content from specific URLs focused on a given research objective.
4. 4Initiates async task runs for longer research jobs, with polling endpoints to retrieve results when complete.

When to use

• You want a drop-in replacement for OpenAI chat that automatically grounds answers in live web data.
• You need to run a large-scale 'find all X matching Y' search against the web asynchronously.
• You're doing deep research on a topic and want cited, sourced answers rather than model hallucinations.

Tools used

Frequently asked questions