Perplexity Search API vs Tavily in 2026

Both calls answered on one key

You do not have to pick one API and rewrite the other. Each call keeps its path, its body and its parser. The address changes, and the key.

# Perplexity: before
curl -X POST https://api.perplexity.ai/search -H "Authorization: Bearer $PERPLEXITY_KEY" -H "Content-Type: application/json" -d '{"query": "open source vector databases", "max_results": 10}'
# Perplexity: after
curl -X POST https://api.specstream.ai/perplexity/search -H "Authorization: Bearer $SPECSTREAM_KEY" -H "Content-Type: application/json" -d '{"query": "open source vector databases", "max_results": 10}'

# Tavily: before
curl -X POST https://api.tavily.com/search -H "Authorization: Bearer $TAVILY_KEY" -H "Content-Type: application/json" -d '{"query": "open source vector databases", "max_results": 10}'
# Tavily: after
curl -X POST https://api.specstream.ai/tavily/search -H "Authorization: Bearer $SPECSTREAM_KEY" -H "Content-Type: application/json" -d '{"query": "open source vector databases", "max_results": 10}'

The parameter and field tables are in the Perplexity Search API guide and the Tavily guide.

What stays the same

On the Perplexity route, query, max_results, search_domain_filter, search_recency_filter, search_after_date_filter, search_before_date_filter, max_tokens and max_tokens_per_page work as on Perplexity, and errors come back in Perplexity's 422 body. On the Tavily route, query, max_results, include_domains, exclude_domains, time_range, start_date, end_date, include_raw_content and include_answer work as on Tavily, and search_depth does not change the price.

Not returned: Perplexity's last_updated, which is always null, and a query list of two or more gets a 422 that asks for one query per call. On Tavily, images is always empty, follow_up_questions is null and results carry no favicon.

One price per search, page text included

Tavily doubles the credits for an advanced search, and Perplexity bills a reply with no results. SpecStream charges $0.005 for a search of up to 10 results on both routes, with the page text, the answer and the advanced depth in that price. There is no monthly reset, credit you buy never expires, and a call stopped by a 429 charges nothing.

The most accurate web search API in the world

The SpecStream Search API 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. The Extract API scores 0.908 F1 on WCXB. Every score and its method is on the benchmarks page.

About one second a search, with no limit per key

Every drop-in route runs the fast search unless the call asks for page text or results past the first ten. On November 9, 2026 production took 1 to 20 fast searches a second 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. There is no limit per key and no daily limit; a busy call gets a 429 with Retry-After and is not charged.

Zero retention on any key

Set zero_retention when you create a key, or later. A search made with that key keeps no request text and no answer after the reply: it is not in the history, its answer is not reused for another call, and its log lines carry neither.

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

A search API returns links and snippets, and your product still has to read the pages and build its records. On SpecStream the same key reads up to 20 pages or documents a call with the Extract API at $0.80 per 1,000 pages, maps and crawls a site at the page price, and runs Deep Search, which returns every match on the live web as structured rows at 5¢, 15¢ or $1 a row. The most a run can cost is approved before it starts. Saved requests rerun hourly to yearly, with each row labeled changed, unchanged or no longer matching, and new rows arrive in Slack, a sheet, a CRM or a signed webhook.

Exa

Search on Exa's own index, at $7 per 1,000 requests on auto and $4 on lite. SpecStream answers Exa's POST /search at https://api.specstream.ai/exa for $5 per 1,000, whatever type you send.

Brave Search API

Search on Brave's own index at $5 per 1,000 requests, with $5 in free credits every month and 50 queries per second. SpecStream answers Brave's call at https://api.specstream.ai/res/v1/web/search with no limit per key.

Serper

Google results sold as prepaid packs, from $1.00 per 1,000 on the 50k pack, with credits valid for 6 months. SpecStream answers Serper's call at https://api.specstream.ai/serper, and its credit never expires.

Parallel

Search, Extract and Task APIs priced by their own units. Web search starts at $1 per 1,000 requests, and the Developer plan allows 600 Search requests a minute. SpecStream answers Parallel's search call at https://api.specstream.ai/parallel with no limit per key.