Category
Search and Retrieval APIs
Search indexes and content APIs that let an agent find, rank, and read web information.
How to read this category
Search APIs are how an agent learns anything it was not told. These three take different positions: Tavily bundles search with extraction, crawling, mapping, and a research endpoint; Exa concentrates on the search call itself with modes from roughly 250 millisecond instant retrieval to deep synthesis runs; Brave rents out an independent index with local and place search on the side.
The practical choice is what happens after the results land. If your agent still needs to read the pages, a bundle or a strong highlights mode saves a second vendor. If you mostly need ranked evidence fast, search depth and latency matter more than bundled extras.
All three are hosted APIs billed by the call. Meter an unattended loop before you trust it; search is cheap per query and expensive per thousand.
The verified profiles
Side by side on sourced facts
| Tool | Pricing model | License | Deployment | MCP support | Checked |
|---|---|---|---|---|---|
| Tavily | Freemium (hosted paid tiers) | Not open source | cloud, API | Official MCP server, local or remote | 2026-10-07 |
| Exa | Paid | Not open source | cloud, API | Official MCP connectivity | 2026-10-07 |
| Brave Search API | Freemium (hosted paid tiers) | Not open source | cloud, API | Official MCP server | 2026-10-07 |
Every cell traces to the tool profile, which traces to the official repository, package page, and documentation checked on the date shown. Pricing cells name the model, not a price: vendor prices change, so check the profile for the sourced pricing summary and the official pricing link.
What actually decides the choice
- Pick by the shape of work: discovery plus reading points at Tavily, latency-tiered search quality points at Exa, an independent index and place data point at Brave.
- Result shape matters: highlights sized for the context window, full text per result, or synthesized answers with sources attached.
- Do not buy a bundle you will not run. Teams that already have extraction covered should compare the search calls alone.
- MCP support exists on all three; check the server pages here for the exact transport and config.
How this archive is ordered, and why it is short
Profiles appear in the order they passed the AgentsUse review gate, not in a tested quality ranking. AgentsUse has not run head-to-head benchmarks for this category, so this page does not claim a winner. The comparison table is the ranking tool: sort by the dimension your job cannot compromise on, then read the full profiles for limitations.
This archive currently lists 3 verified profiles. AgentsUse normally asks for eight published profiles before opening a category archive. This one opens early, on the record, because every listed profile passed the full tool gate and readers comparing real choices need the side-by-side surface now, not a longer unvetted list later. The archive grows as more tools pass review. See how verification works.
Read before you give these tools real work
Guide
How to Evaluate Agent Output Quality
It seems to work is not an evaluation strategy. How to define quality concretely, build runnable evals, and track agent performance over time.
2026-09-29
Guide
10 Tasks AI Agents Handle Well (and 5 to Keep in Your Hands)
A practical look at where AI agents genuinely help in 2026 and where they still fall short, so you can delegate with confidence.
2026-09-26
Back to the tools directory, the category index, or the guides.