Compared with the Perplexity Search API
The Perplexity alternative for news search
Perplexity’s Search API is $5 per 1,000 requests; our normal mode, which reads the top results before ranking them, is $2.20. For plain headline search its Fast Search, at $1, costs the same as our ultra and less than our fast. Same Perplexity SDK, plus a calibrated score, verification and structure on every result.
The same work, mode by mode
Ultra search · headlines only
$1.00
Perplexity: $1Same price
Search API with Fast Search
Fast search
$1.40
Perplexity: $1+40%
Search API with Fast Search
Normal search · reads the top 4
$2.20
Perplexity: $5−56%
Search API (standard)
Contents · per 1k pages
$0.20
Perplexity: $0.50−60%
fetch_url tool (Agent API), per call, plus model tokens
List prices per 1,000 requests (contents, per 1,000 pages), against the cheapest product Perplexity sells for the same work.
Checked September 25, 2026 · Sources: docs.perplexity.ai/getting-started/pricing
What stays the same
Your SDK, your parameters, your response parsing. Here is how each Perplexity parameter maps.
| Perplexity parameter | Status | On typesearch |
|---|---|---|
| query (one, or up to 5) | Same | Same. Several queries are judged together; results come in one list. |
| max_results | Same | Same, 1–20. |
| search_domain_filter | Same | Same: include, or exclude with a “-” prefix. Paths work too. |
| search_recency_filter | Same | hour, day, week, month, year. |
| search_after_date_filter, search_before_date_filter | Same | Same format (MM/DD/YYYY). |
| country, search_language_filter | Mapped | country → only outlets from that country (Perplexity treats it as a preferred region; here it filters) · search_language_filter → languages. |
| max_tokens, max_tokens_per_page | Limited | Ignored: excerpts are always short. |
| search_type: "web", "fast" | Mapped | Both run in our fast mode. For the price of Fast Search, send "mode": "ultra" in the typesearch extras: headlines only. |
| search_type: "people" | Not supported | Returns an error: we index news and the web, not people. |
| Sonar and the Agent API (answers) | Not supported | We don’t write answers. Use essentials for the key facts. |
What you gain
- A calibrated relevance probability on every result
- Top results read and verified before they’re ranked
- Typed questions answered per result, in the same call
- Tone, same-story dedupe and who-published-first
- Similar and contents endpoints with the same key
- Hard cost caps per request
When Perplexity is still the better fit
We’d rather you stay than churn in a month. Keep Perplexity for these:
- You mostly run plain headline searches, or several queries per request: Perplexity’s Fast Search and per-request billing cost less there.
- You need results from every country and language today.
- You want the API itself to write cited answers (its Agent API).
- You need people or finance search.
Migration
The whole migration
One compatibility layer per provider: same parameters, same response shape, same errors. Point your official SDK at us and ship.
from perplexity import Perplexity-client = Perplexity()+client = Perplexity(api_key=TYPESEARCH_API_KEY,+ base_url="https://api.typesearch.ai/compat/perplexity")r = client.search.create(query=["Argentina inflation outlook", "IMF program review"],max_results=10,search_recency_filter="week",)
Switching from Perplexity
What is the best Perplexity alternative for news search?
For agents that search news, typesearch returns results judged for your query, with a calibrated relevance probability on every one and, in normal and deep, the top ones read and verified before they’re ranked. Normal costs $2.20 per 1,000 against $5 for Perplexity’s standard Search API. For plain headline search, Perplexity’s Fast Search ($1) costs the same as our ultra ($1.00, headlines only) for a single query and less than our fast ($1.40); it bills per request of up to five queries, we bill each query.
Can I switch from Perplexity without rewriting my code?
Yes. Keep the Perplexity SDK and point it to https://api.typesearch.ai/compat/perplexity with a typesearch key: the endpoint, parameters and response shape stay the same. The migration is one line, and requests run in our fast mode unless you ask for another one.
Is typesearch cheaper than Perplexity?
Ultra search · headlines only: $1.00, the same as Perplexity ($1, Search API with Fast Search). Fast search: $1.40 against $1 (Search API with Fast Search); Perplexity costs less here. Normal search · reads the top 4: $2.20 against $5 (Search API (standard)). Contents · per 1k pages: $0.20 against $0.50 (fetch_url tool (Agent API), per call, plus model tokens). Perplexity bills per request, and a request can carry up to five queries; we bill each query as a search of its mode. So for plain headline search, or several queries per request, Perplexity can cost less; for news read and verified before it is ranked, our normal mode costs less than its standard Search API.
When should I stay with Perplexity?
If you need answers written with citations (its Agent API, which replaces Sonar on September 27, 2026), people or finance search, plain headline search billed per request of up to five queries (Fast Search, $1 per 1,000 requests), or results from every country and language. Since the SDK is the same, you can also use both: send each request where it fits.
Switch from Perplexity
Free key, same SDK, 2.2× cheaper in normal mode.