Jina Reader (r.jina.ai) compatible

The call, before and after

# Before
curl -X POST https://r.jina.ai/ -H "Authorization: Bearer $JINA_API_KEY" -d "url=$URL"

# After
curl -X POST https://api.specstream.ai/jina -H "Authorization: Bearer $SPECSTREAM_API_KEY" -d "url=$URL"

The body, the headers and the parser stay. The reply starts with Title: and URL Source:, then Markdown Content:, as Jina's does. The address form https://r.jina.ai/<url> is not served: send the page as url in the body.

What the route reads, and what it refuses

Read and applied: url in a form body or a JSON body, Accept: application/json for the JSON reply, and X-Return-Format set to markdown or text. With text the JSON reply carries data.text in place of data.content.

Refused in Jina's own form: no Authorization header gets 401 AuthenticationFailedError, no url gets 400 ParamValidationError, and a host that does not resolve gets 422 SubmittedDataMalformedError. Every call was measured on production on Nov 10, 2026, and the full pass table lists each parameter and its result.

Every field that comes back

The JSON reply has code 200, status 20000 and data with title, url, content and publishedTime when the page has a date. The route does not return data.description, links, images, warning, metadata or usage.tokens.

One field is added: specstream, with the call's price_usd and request ID. A parser that reads Jina's fields never sees it.

$0.80 per 1,000 pages, in dollars

Jina charges tokens by the length of each page, so a long page costs more than a short one. On SpecStream every page read costs $0.0008, whatever its length, and a page that gives no text is not charged. A busy call gets 429 with Retry-After and charges nothing, and credit you buy never expires.

0.908 F1 on WCXB, and the search scores on the same key

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

You decide what is kept, on each key

Create the key with zero_retention and a page read made with it keeps no request text and no answer after the reply. 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 searches the live web with the Search API at $5 per 1,000 searches of up to 10 results. 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. The most a run can cost is approved before it starts. The full comparison is on the Jina Reader alternative page.

Firecrawl

A scrape API at 1 credit per page, with 1,000 free credits a month and the Hobby plan at $19 a month, in its own request and reply. SpecStream reads a page for $0.0008 and answers the Jina call your code already sends.

Exa Contents

Page text from Exa at $1 per 1,000 pages for each content type you ask for, in Exa's request and reply. SpecStream reads 1,000 pages for $0.80 and keeps Jina's reply shape.

Tavily Extract

Every 5 successful URL extractions cost 1 credit, at $0.008 a credit on pay as you go, in Tavily's own reply. SpecStream charges $0.80 per 1,000 pages with no credits and no monthly reset.