Compared with mediastack
The mediastack alternative for AI agents
mediastack is a simple, low-cost feed of article metadata. typesearch adds what an agent needs on top: a calibrated relevance probability on every result, the top ones read before they’re ranked, typed answers, tone and same-story grouping — pay as you go, from $1.40 per 1,000 searches.
Checked Sep 25, 2026 · A simple, low-cost news API
At the same volume
Standard · 10,000 requests a month
$14/ month
typesearch, fast mode
mediastack: $24.99−43%
$2.50 per 1,000
Professional · 50,000 requests a month
$70/ month
typesearch, fast mode
mediastack: $99.99−29%
$2 per 1,000
Business · 250,000 requests a month
$350/ month
typesearch, fast mode
mediastack: $249.9929% less on mediastack
$1 per 1,000
One call returns up to 100 articles on mediastack and up to 50 judged results on typesearch. Overage on mediastack is billed per call above the plan; annual billing is up to 15% less. typesearch has no plan: the same price at any volume.
Checked Sep 25, 2026 · Sources: mediastack.com/pricing
Side by side
What mediastack publishes, next to what typesearch does. Each of their facts links to where we read it.
Checked Sep 25, 2026 · Sources: mediastack.com/pricing, docs.apilayer.com/mediastack/docs/api-features, docs.apilayer.com/mediastack/docs/getting-started, docs.apilayer.com/mediastack/docs/code-examples
How the parameters map
Each mediastack parameter and what it becomes on typesearch. You rewrite one request; your agent’s logic stays.
| mediastack parameter | Status | On typesearch |
|---|---|---|
| keywords | Mapped | query. Or up to five queries, judged together. |
| sources (and -source) | Mapped | include_domains, exclude_domains, up to 20 each. |
| categories | Mapped | sections: the outlet’s own section, or the start of the URL path. |
| countries, languages | Limited | Query in the language you want, or restrict to regional outlets with include_domains. |
| date | Mapped | published_after, published_before, or days, within the last 30 days. |
| sort | Limited | Ranked by a calibrated relevance probability; published_at to sort by date. |
| limit, offset | Limited | max_results, up to 50. No pagination. |
| access_key (in the URL) | Mapped | Authorization: Bearer <key>, in a header |
| data[]: title, url, description, published_at, source | Mapped | results[]: title, url, snippet, published_at, source |
What you gain
- A calibrated relevance probability on every result, instead of date or popularity
- Top results read and verified before they’re ranked (normal and deep)
- Typed questions, tone and same-story grouping in the same call
- SDK compatibility with Exa, Tako and Perplexity, and an OpenAPI 3.1 contract
- The key travels in a header, never in the URL
When mediastack is the better fit
We’d rather you pick right the first time. Keep mediastack for these:
- You need up to 100 articles per call and filter them yourself.
- You use most of the Business plan: at 250,000 calls a month it costs less per call than our fast mode.
- You filter by country and language across 50+ countries today.
The migration
The whole migration
Before and after: the request, the key header and the fields you read. Everything else in your code stays.
import requests-r = requests.get("https://api.mediastack.com/v1/news",- params={"access_key": MEDIASTACK_KEY, "keywords": "lithium royalties in Chile", "date": "2026-09-18,2026-09-25", "limit": 20},-).json()+r = requests.post("https://api.typesearch.ai/v1/search",+ headers={"Authorization": f"Bearer {TYPESEARCH_API_KEY}"},+ json={"query": "lithium royalties in Chile", "days": 7, "max_results": 20, "mode": "fast"},+).json()-for a in r["data"]:- print(a["title"], a["url"], a["published_at"], a["source"])+for a in r["results"]:+ print(a["title"], a["url"], a["published_at"], a["source"], a["score"])
Switching from mediastack
How is typesearch different from mediastack?
mediastack returns article metadata filtered by keyword, source, country, language and category, sorted by date or popularity. typesearch judges every result for your query and adds typed answers, tone and same-story grouping, with the key in a header instead of the URL.
How much does typesearch cost compared with mediastack?
At the volume of mediastack’s Standard plan (10,000 requests a month), typesearch costs $14 in fast mode, against $24.99. At the volume of mediastack’s Professional plan (50,000 requests a month), typesearch costs $70 in fast mode, against $99.99. At the volume of mediastack’s Business plan (250,000 requests a month), typesearch costs $350 in fast mode, against $249.99. Results served from the cache are free, and there are no seats or minimums.
How do I switch from mediastack?
Rewrite one request: the endpoint, the key header and a few parameter names change — see the table above. The fields you read map one to one: description → snippet, published_at → published_at, source → source.
When should I stay with mediastack?
If you need up to 100 articles per call, or the lowest price per call at 250,000 calls a month. You can also use both: typesearch for what your agent searches now, mediastack for the rest.
Keep comparing
Other comparisons
Use cases
Switch from mediastack
A key in a minute, $5 in free credit, one request to rewrite.