Web search for your AI app

The app sends its SearXNG call, and SpecStream answers it

This is the call LibreChat, GPT Researcher, n8n, Agno and the others send once the address is set. Try it with your key before you change the app.

curl "https://api.specstream.ai/searxng/$SPECSTREAM_API_KEY/search?q=who+founded+Mistral+AI&format=json"

The reply is SearXNG's JSON: results[] with title, url and content, the fields these apps read. Each call is one search of 10 results at $0.005.

No instance to run

A SearXNG you host needs json under search.formats in its settings.yml before an app can read it, and a server to keep running. The SpecStream address answers JSON from the first call and searches the whole live web at the second of the call, with nothing to install.

The key travels in the address, so apps that send no key for SearXNG still sign their calls. It also goes in key=, an X-API-Key header or a Bearer token where the app has a key field.

$0.005 a search, and a failed call costs nothing

A search of up to 10 results costs $0.005, which is $5 per 1,000. Each started block of 10 more results adds $0.006, up to 100 results a search. A call stopped by a 429 charges nothing, and a top-up never expires.

A new account gets $10 of credit with no card, which covers 2,000 searches.

The most accurate web search API, at about one second a call

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

A drop-in call for the first ten results, without page text or the answer, runs the fast search. On November 9, 2026 production answered 718 of 720 searches at 1 to 20 a second, with a median of 0.82 s at 1 a second and 1.07 s at 20 a second, and the 95th percentile between 1.09 and 1.73 s. One account takes up to 8,800 searches a minute; a call over that gets a 429 with Retry-After and is not charged.

No query text is kept on a zero retention key

Create a key with zero_retention and a search made with it keeps no request text and no answer after the reply, in the history or in a log line. 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}'

Pages and finished rows on the same 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 returns every match for a request in plain words as structured rows, at 5¢, 15¢ or $1 a row, with a source URL on every value.

Tavily

Offered in Open WebUI, LibreChat, AnythingLLM and GPT Researcher. $0.008 a credit on pay as you go, one credit for a basic search and two for an advanced one, with 1,000 free credits a month; when they run out, requests stop until the reset. SpecStream charges $0.005 a search, and a top-up never expires.

Brave Search API

Offered in Open WebUI, AnythingLLM, Dify and n8n. $5 per 1,000 requests, with $5 in free credits every month. SpecStream costs the same $5 per 1,000 and scores 97.9 F-score on all 1,000 SimpleQA Verified questions.

Exa

Offered in Open WebUI and AnythingLLM. $4 per 1,000 requests on lite and $7 per 1,000 on auto and fast. SpecStream is $5 per 1,000 for every search, at about one second a call.

Serper

Offered in Open WebUI, LibreChat, AnythingLLM, GPT Researcher and Dify. Credit packs start at 50,000 credits for $1.00 per 1,000, and credits are valid for 6 months. A SpecStream top-up never expires.

Firecrawl

Offered in Open WebUI, LibreChat and Dify. A search costs 2 credits per 10 results, and the Hobby plan is $19 a month. On SpecStream a search is $0.005 with no plan to buy, and a page read is $0.0008.