Vertex AI Search vs Custom Search JSON API

One line in place of a new integration

Moving to Vertex AI Search means a Google Cloud project, an engine, a serving config and a new endpoint. Moving to SpecStream means a new host and a new key.

# Before
curl "https://www.googleapis.com/customsearch/v1?key=$GOOGLE_KEY&cx=$CX&q=open+source+vector+databases"

# After
curl "https://api.specstream.ai/customsearch/v1?key=$SPECSTREAM_KEY&cx=$CX&q=open+source+vector+databases"

The reply has Google's shape, field for field: kind, queries with request and nextPage, searchInformation and items.

The whole live web from the first call

Google's page names Vertex AI Search for searching up to 50 domains and asks you to contact Google for full web search. SpecStream searches the whole live web on every call. Send siteSearch when you want one site.

What stays the same

q, num (1 to 10) and start work as before, and queries.nextPage gives the next start. exactTerms, orTerms, siteSearch and hq shape the search. cx is accepted and not used. The key goes in key=, in X-Goog-Api-Key or as a Bearer token.

One price per search, nothing for storage

Vertex AI Search bills search queries and data storage. SpecStream bills the search: $5 per 1,000 searches of up to 10 results, and each started block of 10 more results adds $0.006. A call stopped by a 429 charges nothing, and credit you top up never expires.

The most accurate web search API in the world

On SimpleQA Verified the SpecStream Search API scores 97.9 F-score on all 1,000 official questions, 978 of them correct. On WideSearch it scores 64.0 item F1 and 43.4 row F1 on all 200 official tasks. The Extract API scores 0.908 F1 on WCXB. Every score and its method is on the benchmarks page.

About one second a search, measured under load

The Custom Search route runs on the fast search. On November 9, 2026 production took 1, 5, 10 and 20 searches a second through the fast search and answered 718 of 720. The median was 0.82 s at 1 search a second and 1.07 s at 20 a second, and the 95th percentile stayed between 1.09 and 1.73 s. One account can send up to 8,800 searches a minute; a call over that gets a 429 with Retry-After and is not charged.

You decide what is kept, on each key

Create a key with zero_retention and a search made with it keeps no request text and no answer after the reply. It is not in the history and its log lines carry neither. The charge keeps the account, the key, the request ID, the price and the time. spend_cap_cents is the most the key can spend.

curl -X POST https://api.specstream.ai/v1/account/keys \
  -H "Authorization: Bearer $SPECSTREAM_API_KEY" -H "Content-Type: application/json" \
  -d '{"name": "production", "scopes": ["account:read", "finds:spend"], "spend_cap_cents": 500000, "zero_retention": true}'

Search, pages and rows on one key

The same key reads up to 20 pages or documents a call with the Extract API at $0.80 per 1,000 pages, and a page with no text is not charged. Deep Search takes a request in plain words and returns every match on the live web as structured rows, at 5¢, 15¢ or $1 a row, plus 5¢ for an email and 15¢ for a phone. The most a run can cost is quoted before it starts, and a saved request reruns hourly to yearly.

Grounding with Google Search

A Gemini API tool at $14 per 1,000 search requests on Gemini 3 and newer models. It returns the model's grounded answer, not a result list. SpecStream returns the list in Google's JSON at $5 per 1,000. See the comparison.

Programmable Search Element

The search box Google keeps for your own websites, free for up to 50 domains. SpecStream answers your code's JSON call across the whole live web.

Brave Search API

$5 per 1,000 requests, up to 50 queries a second, with $5 in free credits every month. Brave's own guide says moving to it means updating how you read the response. On SpecStream your parser stays.

Exa

$4 per 1,000 requests for lite search and $7 for auto and fast, with its own request and reply. SpecStream answers the Custom Search call you send today at $5 per 1,000.