Web search and reading for XIA

Ask XIA to search the web.

XIA follows what matters to you: it searches, reads pages and PDFs, and brings back content that's clean and ready to act on. Nothing to install, no own APIs, no connectors to maintain.

Try it freeHow it works
PDFs includedPage cacheQuota per planPublic API
road signage public tenders in Mendoza 2026

Public Tender N° 12/2026 — Road signage

gob.mendoza.gov.ar/licitaciones/12-2026

Construction and installation of road signage on urban routes…

Terms and conditions — Signage

transport.mendoza.gov.ar/pliegos/12-2026.pdf

PDF · 84 pages · downloadable bid requirements

2025 awards — provincial public works

datos.mendoza.gov.ar/adjudicaciones-2025

Full list of awardees and amounts for the period…

Search: the agent asks and answers with sources in seconds.

pliego-signage-12-2026.pdf

Requirements detected

Bid guarantee5% of official budget
Delivery deadline45 calendar days
Track recordSimilar works ≥ 20M in 5 years

PDF reading: an 84-page bid document becomes a table.

curl -X POST https://www.napsix.com/api/gateway/v1/search \
  -H "Authorization: Bearer $XIA_API_KEY" \
  -d {"q": "steel pipe suppliers in Rosario", "max": 10}
{
  "results": [
    { "title": "…", "url": "…", "snippet": "…" }
  ]
}

The same power from your code: POST /v1/search.

How it works

Three steps, one agent that delivers.

XIA picks the path: if the result is a PDF it reads it directly, and if a site blocks or asks for a login, it knows and tells you.

  • The agent searches

    web_search: up to ten results with title, URL and summary, in a couple of seconds.

  • It reads the page or the PDF

    scrape_url returns clean Markdown; PDFs — bid documents, catalogs, spec sheets — at no extra cost.

  • It acts on what it found

    A report, a comparison sheet or CRM contacts: the next turn already uses the data.

What it's for

Work that used to be hours of open tabs.

The most common cases among teams using the platform.

  • Scan tenders and bid documents

    Long PDFs full of requirements: the agent reads them and brings you the table you need.

  • Map suppliers

    Catalogs, spec sheets and plants by region, ready to compare.

  • Track competitor prices

    Monitor competitors' pricing and get the updated figure on each pass.

  • Know a client before the meeting

    Read the client's web: their history, products and team, in one turn.

  • Stay current in your industry

    Sector news and updates, summarized daily.

  • Feed your Knowledge Base

    Your own web, sitemap or RSS get indexed and become searchable for the whole team.

Prospecting

Find and enrich leads, for real.

You set the intent; XIA searches, enriches and leaves the work ready to review.

  • Set the intent

    An industry and a region — or a phrase: “food shops in Córdoba”.

  • XIA searches and enriches

    Web and public sources: contact, socials, location, sector. No bought lists.

  • It lands in a sheet

    You review and edit everything before it touches your CRM.

  • One click, and it's in your CRM

    Contacts arrive with their story; the pipeline starts with context.

Try the flowSee XIA Agents
Napsix Knowledge Base

The KB feeds itself.

Four kinds of web sources: set them up once and XIA keeps them fresh. Documents become searchable for your team and your agents.

  • URL

    A single page: index it and it's in the base.

  • Sitemap

    The whole site from one source, refreshed automatically.

  • RSS

    Posts and news: every new article just lands.

  • Full site

    A crawl that respects robots.txt, with visible status.

See NapsixStorage
Napsix AI Gateway

Your code uses the rail too.

Two endpoints in the Gateway, with your same nsx_live_* key: scrape returns clean Markdown, HTML or links — and PDFs go straight through, no extra steps.

curl -X POST https://www.napsix.com/api/gateway/v1/search \
  -H "Authorization: Bearer nsx_live_*" \
  -d {"q": "steel pipe suppliers in Rosario", "max": 10}
curl -X POST https://www.napsix.com/api/gateway/v1/scrape \
  -H "Authorization: Bearer nsx_live_*" \
  -d {"url": "https://example.com/bid-doc.pdf", "format": "markdown"}

Daily quota of the Gateway plan

50
Free
1,000
Starter
10,000
Pro
Unlimited
Enterprise

Every call, success or not, is audited in your Gateway usage history.

Explore the GatewayDeveloper docs
Included limits

Every plan with its daily quota.

Searches and reads per day, according to your XIA plan. They reset every day.

CriterioFreeStarterProBusinessEnterprise
Web searches / day201506002,000Unlimited
Reads / day201506002,000Unlimited
PDFs in readsSíSíSíSíSí
Page cacheSíSíSíSíSí

The Gateway API also has its own per-plan quota.

See all plans
FAQ

Common questions

What happens if a site blocks XIA or asks for a login?
The agent knows: it won't serve the error page as if it were content. It tells you in the chat and continues with what it has.
Does it read PDFs?
Yes, natively and at no extra cost: bid documents, catalogs, spec sheets, reports.
Does it respect robots.txt?
Yes. Full-site crawls respect robots.txt, and a blocked or broken page is never presented as content.
What counts as a read?
A page or PDF read. If the page is already cached, it comes back in milliseconds without using quota.
Can I use it from my own code?
Yes: POST /v1/search and POST /v1/scrape in the Napsix AI Gateway, with your nsx_live_* key.
What happens when the daily quota runs out?
The agent tells you and continues with what it already has. The quota resets every day.

Have your agents read the web.

Start free: the Free tier includes 20 searches and 20 reads per day.

Sign upSee plans

Works with

XIA AgentsNapsixStorageAI Gateway
XIA
XIAby Napsix.AI
public