Give your agents the news the moment it breaks. Built for trading bots, risk monitors and research copilots that must act on what happened minutes ago, in any language.
Your agent is only as smart as its last update. typesearch feeds it fresh, verified news from newsrooms worldwide, so it decides on what’s happening now, not on last year’s training data.
From $1.00 per 1,000 searches · Exa and Tako list search at $7
Sample outputRelevance is a calibrated probability: filter by threshold and keep only what matters.
Speaks the SDKs you already ship
Quickstart
From sign-up to ranked results in a couple of minutes, with the tools you already use.
TYPESEARCH_API_KEYpip install typesearch1from typesearch import Typesearch23ts = Typesearch() # TYPESEARCH_API_KEY45res = ts.search(6 "Fed rate cut expectations",7 mode="fast",8 max_results=5,9)1011for r in res.results:12 print(r.score, r.title)
The numbers
Public list prices and our own measurements. No cherry-picked benchmarks.
| typesearch | Exa | Tako | |
|---|---|---|---|
| Ultra search · headlines only | $1.00 | $7 | $7 |
| Fast search | $1.40 | $7 | $7 |
| Normal search · reads the top 4 | $2.20 | $7 | $11 |
| Deep search · reads 8 | $5.60 | $12–15 | $20 |
| Contents · per 1k pages | $0.20 | $1 | $1 |
Prices: exa.ai/pricing and tako.com/pricing, September 25, 2026. Speed and compatibility: our internal test suite. Methodology on request.
Search API
Not an opaque rank: a probability you can act on. Set a threshold, say 0.8, and your agent works only with what’s truly relevant. No guesswork, no extra filtering step.
Savings accounts: what changes for depositors
Rate decision: what it means for mortgages
Headlines promise; the article decides. In Normal and Deep we open the top candidates and check them against your query, so only results that hold up reach your agent.
// Mentions a specific figure? "figure": { probability: 0.93 }, // Main actor "actor": { choice: "central bank", confidence: 0.88 }, // Impact on consumers (1–5) "impact": { score: 3.6, confidence: 0.81 }
Ask questions alongside the search (yes/no, multiple choice, a 1–5 score) and every result comes back already answered. One call, no second LLM pass, no parsing.
Charts APINew
Ask in plain words. We find the figures in the news, pick the right chart and write the headline. Every value links to its source.
Try a question
From $1.50 per 1,000 charts · Tako lists at $7
Sample card
Do the math
Pick a mode and your monthly volume, and see what you’d pay here versus Exa, Tako and Perplexity at list prices.
Public list prices from exa.ai/pricing, tako.com/pricing and docs.perplexity.ai, checked September 25, 2026, before free tiers. Each mode is compared with the closest product each one sells: ultra and fast with a search (Perplexity: Fast Search; Tako: Instant or Fast); normal with a search that returns page text (Exa includes it, Tako adds 4 contents, Perplexity’s standard Search API); deep with Exa’s Deep Search or Tako’s Deep plus 8 contents (Perplexity has no equivalent); contents per page. The comparison is against the cheapest of the three.
You keep
$33,600a year
The index
We crawl and rank the news ourselves, not through a reseller. The busiest sources are re-read every few minutes, so new stories are searchable right after they’re published.
Beyond ten blue links
Whether each result is positive, neutral or negative about what you asked, not about the world in general. Per result and per source.
The same wire story published by ten outlets comes back as one result, with the other nine attached.
See how a story spread: coverage by day and by outlet, and who broke it first. Included at no extra cost.
Up to three verbatim quotes from different sources. We quote, we never write, so there’s nothing to hallucinate.
Set a maximum spend per request. If it’s reached, you get what fits and the response is flagged incomplete. No surprise invoices.
Search any site in real time: we read its homepage, key sections and its own search box, in parallel.
Migration
One compatibility layer per provider: same parameters, same response shape, same errors. Point your official SDK at us and ship.
from exa_py import Exa-exa = Exa(api_key=EXA_API_KEY)+exa = Exa(api_key=TYPESEARCH_API_KEY, base_url="https://api.typesearch.ai/compat/exa")results = exa.search("lithium royalties in Chile", num_results=10, type="fast")similar = exa.find_similar(results.results[0].url, exclude_source_domain=True)pages = exa.get_contents([r.url for r in results.results], highlights=True)
Production
/v1 only changes by adding. Anything breaking ships as /v2, with notice.
Daily quotas and per-minute limits on every key, visible in real time through the usage API.
RFC 9457 problem details, stable error codes and a request ID on every response.
We store a hash, never your key. Rotate and revoke anytime.
The contract is generated from the same schemas that validate every request.
JSON or server-sent events on every search endpoint, with partial results as they land.
Volume pricing, custom source coverage, dedicated support and SLAs. Tell us what you pay today — we’ll show you what you’d pay here.
Yes. Every new account gets $5 in free credit — about 5,000 ultra searches, 3,571 fast or 2,272 normal ones — valid for 30 days. No card needed. Your first top-up gets 20% extra, up to $20.
It depends on the mode and on who you compare with. An ultra search, which judges headlines only, lists at $1.00 per 1,000 and a fast one at $1.40, against $7 for a search on Exa or Tako: up to 7× less. Normal and deep read pages before ranking, so they cost more ($2.20 and $5.60), and since March 2026 Exa’s search includes the text of its results, so the gap in those modes is smaller. Some APIs, such as Perplexity’s and Parallel’s, list their most basic search at the price of our ultra mode and for less than our fast mode. The pricing page compares every mode; put your own volume in the calculator above.
It shouldn’t. We run the official SDKs — Python and JavaScript — against our compatibility layers: 25 of 25 test cases pass by changing only the base URL and the key. Anything we don’t support is accepted and reported in a response header, not thrown as an error.
News and editorial content, crawled and refreshed by us — the busiest sources every few minutes — focused on local and regional press, starting with Latin America, from 130+ countries in 30+ languages. We don’t publish the list of sources: every result shows its own and links to the original. Missing an outlet? Suggest it from the dashboard (Suggest a source) or write to support@typesearch.ai.
No, on purpose. We return evidence: ranked, verified, with verbatim excerpts. If you need prose, your model writes it — grounded in results that can’t be hallucinated.
If your agent works with news, current events or anything being published right now, yes: that’s exactly what we’re built for. For people, company or academic-paper search, pair us with a specialized index; with the same SDK, each request goes where it fits best.
Get a key in a minute, point your SDK at us and watch your bill shrink. $5 free to start, no card.