search.semantic.web
Find conceptually related web content using semantic search with filters for categories, domains, and dates to retrieve relevant pages with highlighted excerpts.
Instructions
Neural/semantic web search — finds conceptually related pages, not just keyword matches. Supports category filters (company, research paper, news, people, tweet), domain filtering, date range. Returns relevance scores and highlighted excerpts (Exa)
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes | Natural language search query — Exa finds semantically related pages, not just keyword matches | |
| num_results | No | Number of results (default 10, max 25) | |
| type | No | Search type: "auto" (balanced), "neural" (semantic similarity), "keyword" (traditional). Default: auto | |
| category | No | Category filter for specialized indexes (optional) | |
| include_domains | No | Only include results from these domains (e.g. ["arxiv.org"]) | |
| exclude_domains | No | Exclude results from these domains | |
| start_published_date | No | Filter: only results published after this date (ISO format, e.g. "2025-01-01") | |
| end_published_date | No | Filter: only results published before this date | |
| include_text | No | Include full extracted page text in results (default false — saves tokens) | |
| include_highlights | No | Include key sentence highlights (default true) |