Skip to the page
Get started free

Telnyx Web Search API

Search at $1 per 1,000, page text included.

Telnyx launched its Web Search API on August 12, 2026. Search, page content and cited research each cost $5 per 1,000 calls, billed through your Telnyx account, at 60 requests a minute per key by default.

SpecStream is the live web data engine. Its Search API costs $1 per 1,000 searches on the fast tier, with the text of each page closest to your query in the same reply, and its Extract API $0.80 per 1,000 pages.

It is the most accurate web search API in the world: 97.9 F-score on all 1,000 SimpleQA Verified questions (978 correct) and 64.0 item F1 on all 200 WideSearch tasks, on the benchmarks page.

Get an API key

New accounts get $10 of credit, no card. It covers 10,000 fast searches.

SpecStream and Telnyx Web Search side by side

SpecStreamTelnyx Web Search
The callPOST https://api.specstream.ai/v1/web-search with {"q": ..., "speed": "fast"} and a SpecStream key as a Bearer token, or GET with the same parameters in the addressPOST https://api.telnyx.com/v2/web_search with query and count (1 to 100), and a Telnyx key as a Bearer token
Price per 1,000 searches$1 for up to 10 results on the fast tier. $5 for the default search, which adds the answer and typed rows$5 per 1,000 calls
Page text in the searchAdd passages=true and each source carries excerpt: the text of its page closest to the query, up to about 3,500 characters, read liveSnippets averaged 322 characters in Telnyx's own 20 query test. The page text is a second call to the contents endpoint
Reading pagesExtract API: up to 20 URLs a call at $0.80 per 1,000 pages, PDF, Word and Excel included. A page with no text is not charged, and a page that refuses a plain read is tried in a real browser, then from the site's own feed, then from its newest public archive captureContents: up to 20 URLs a call at $5 per 1,000 calls, served from a crawl cache when possible
A cited answerThe default search returns answer with the page it was read from and its read time, every source read, and rows, in a few seconds, at $5 per 1,000Research at $5 per 1,000 calls: answer and citations, in tens of seconds and up to 120, at lite, standard or deep effort
RateNo limit per key and no daily limit on the fast search. A call that has to wait gets Retry-After and is not charged60 requests a minute per API key by default
SpeedMedian 0.82 s at one search a second and 1.07 s at 20 a second on the fast search, measured on production on November 9, 20261,027 ms on average over 20 queries, in Telnyx's own test
What answers the searchThe whole live web, read at the moment of the call. 97.9 F-score on all 1,000 SimpleQA Verified questions and 64.0 item F1 on all 200 WideSearch tasksA provider behind Telnyx's own interface. Telnyx says the backend is swappable without client side updates
Account and billing$10 of credit on sign-up, no card. A spending limit on each key, and X-End-User-Cap-Cents caps what each of your own users spends in a monthBilled through your Telnyx account, on the key you use for voice, messaging and inference
Many results as rowsDeep Search: a request in plain words returns every match on the live web as rows that meet each condition you wrote, at 5¢, 15¢ or $1 a row. The most a run can cost is quoted before it startsResearch returns answer and citations

Why teams choose SpecStream over Telnyx Web Search

Your Telnyx search, on SpecStream

The request keeps its words. The address, the key and two field names change.

# Telnyx: $5 per 1,000 calls, snippets
curl -X POST https://api.telnyx.com/v2/web_search \
  -H "Authorization: Bearer $TELNYX_API_KEY" -H "Content-Type: application/json" \
  -d '{"query": "latest voice AI frameworks", "count": 5, "country": "US"}'

# SpecStream: $1 per 1,000 searches, page text in the same reply
curl -X POST https://api.specstream.ai/v1/web-search \
  -H "Authorization: Bearer $SPECSTREAM_API_KEY" -H "Content-Type: application/json" \
  -d '{"q": "latest voice AI frameworks", "num": 5, "country": "US", "speed": "fast", "passages": true}'

query is q, count is num (1 to 100), country stays country and safesearch is safe. include_domains becomes site: in q and exclude_domains becomes -site:, and both can be in one query. freshness becomes after: with a date. The results are in data.sources[], each with url, title, snippet and, with passages, its excerpt.

One call where Telnyx takes two

On Telnyx a search returns snippets, and the agent calls the contents endpoint for the pages it picked: a search and one contents call cost $0.01 together. On SpecStream, passages=true puts the text of each page closest to your query into the search reply, for $0.001 a search of up to 10 results.

Pages read live, documents included

The Extract API reads up to 20 URLs a call, web pages and PDF, Word and Excel files, at $0.80 per 1,000 pages. A page with no text is not charged. A page that refuses a plain read is tried as a browser would ask for it, then in a real browser, then from the site's own feed, then from its newest public archive capture, and read_from says which one answered.

A cited answer in a few seconds

Telnyx research answers in tens of seconds, up to 120. The SpecStream default search answers a question from the live web in a few seconds: answer holds the value as the pages print it, the page it was read from and its read time, and a question that asks several things is answered only when each part is. It costs $5 per 1,000.

The most accurate web search API in the world

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

Built for a product that serves its own users

Send X-End-User-Id with each call and usage is summed per user, and X-End-User-Cap-Cents caps what that user spends in the month. Each key carries its own spending limit. A key can be set to zero retention, so no request text or answer is kept after the reply, or to send its searches only from our server in Madrid, in the EU. Every call is in the request log with its price.

From one search to every match as a row

When the job is a dataset, not one answer, the same key runs Deep Search. You write the request in plain words, SpecStream reads it back as the conditions every row must meet and the facts to bring back, and returns every match on the live web as rows at 5¢, 15¢ or $1 a row, each value with its source URL and read time. The most the run can cost is approved before it starts.

Other Telnyx Web Search alternatives

Brave Search API

A search API on Brave's own index at $5 per 1,000 requests. SpecStream answers Brave's /res/v1/web/search call on the same path, at $1 per 1,000 searches of up to 10 results.

Exa

Search at $4 per 1,000 requests on lite and $7 on auto, for up to 10 results. SpecStream answers Exa's POST /search at https://api.specstream.ai/exa for $1 per 1,000, and reads each page live. The Exa comparison has every row.

Tavily

Billed in credits at $0.008 each pay as you go, and a basic search is 1 credit, so $8 per 1,000. SpecStream answers Tavily's call at https://api.specstream.ai/tavily/search for $1 per 1,000.

Perplexity Search API

Ranked results as JSON, and $1 per 1,000 requests with Fast Search. SpecStream answers Perplexity's POST /search at https://api.specstream.ai/perplexity for $1 per 1,000, with the Extract API and Deep Search on the same key.

How SpecStream scores

The same numbers as the benchmarks page, from the official tasks and scorers.

SpecStreamParallelExaTavilyPerplexity
Web researchWideSearch, GPT-5.6 Luna60.945.553.047.947.0
Factual accuracySimpleQA Verified, GPT-5.6 Luna9694919494
Factual accuracySimpleQA Verified, GPT-5.6 Sol9997919295

How to switch from Telnyx Web Search

1. Create a key under Agents & API in the workspace. A new account gets $10 of credit, no card, which covers 10,000 fast searches.

2. Point your search tool at SpecStream. This function returns Telnyx's result shape, so the code that reads title, url, description and snippets stays as it is.

import requests

API = "https://api.specstream.ai/v1/web-search"  # was https://api.telnyx.com/v2/web_search


def web_search(query, count=10, country=None, include_domain=None):
    q = f"{query} site:{include_domain}" if include_domain else query
    body = {"q": q, "num": count, "speed": "fast", "passages": True}
    if country:
        body["country"] = country
    data = requests.post(API, headers={"Authorization": f"Bearer {SPECSTREAM_KEY}"}, json=body).json()["data"]
    return [
        {"url": s["url"], "title": s["title"], "description": s["snippet"], "snippets": [s.get("excerpt") or s["snippet"]]}
        for s in data["sources"]
    ]

3. Move page reads to POST https://api.specstream.ai/v1/extract with {"urls": [...], "format": "markdown"}. Each item of data.results has url, status, title and text.

4. Move research questions to the default search, GET https://api.specstream.ai/v1/web-search?q=<the question>, and read data.answer and data.sources.

Get an API key

Frequently asked questions

Something we didn't answer? Talk to a human

How much does the Telnyx Web Search API cost?
$5 per 1,000 calls for each operation: web search, content retrieval and deep research, billed through your Telnyx account. A search and one contents call for its pages cost $0.01 together. A SpecStream fast search costs $0.001, with the page text in the same reply.
Do I need a Telnyx account to use it?
Yes. Every endpoint takes the Telnyx key you use for voice, messaging and inference, and the calls are billed through your Telnyx account. SpecStream needs only a SpecStream key: sign up, get $10 of credit with no card, and create a key under Agents & API.
What is the Telnyx Web Search rate limit?
60 requests a minute per API key by default, by Telnyx's API reference. SpecStream's fast search has no limit per key and no daily limit, and a call that has to wait gets Retry-After and is not charged.
Which search provider is behind Telnyx Web Search?
Telnyx's product page says Telnyx owns the API surface and the backend is swappable without client side updates. SpecStream searches the whole live web and reads each page at the moment of the call, and scores 97.9 F-score on all 1,000 SimpleQA Verified questions.
How long does Telnyx research take?
Tens of seconds in practice and up to 120, depending on research_effort, by Telnyx's docs. The SpecStream default search answers a question in a few seconds, with the page the answer was read from and its read time, at $5 per 1,000.
Does SpecStream answer the Telnyx request as it is?
No, the request changes in two fields: query becomes q and count becomes num, at https://api.specstream.ai/v1/web-search. The function above returns Telnyx's result shape, so the rest of your agent stays as it is.
Does SpecStream have an MCP server?
Yes, at https://api.specstream.ai/mcp over Streamable HTTP, with OAuth or an API key. The search_web tool runs a search and read_pages reads pages, from the same balance and inside the limits you set on the key.

Sources

SpecStream wrote this page. Each rival figure comes from that company's own page.

  1. Telnyx Web Search API release notetelnyx.com/release-notes/web-search-api, read Nov 12, 2026
  2. Telnyx Web Search API product pagetelnyx.com/products/web-search-api, read Nov 12, 2026
  3. Telnyx Web Search API referencedevelopers.telnyx.com/api-reference/web-search/web-search.md, read Nov 12, 2026
  4. Telnyx Web Search docsdevelopers.telnyx.com/docs/web-search.md, read Nov 12, 2026
  5. Telnyx search API benchmarktelnyx.com/resources/web-search-api-benchmark, read Nov 12, 2026
  6. Brave Search APIbrave.com/search/api/, read Nov 12, 2026
  7. Exa pricingexa.ai/docs/admin/pricing.md, read Nov 12, 2026
  8. Tavily credits and pricingdocs.tavily.com/documentation/api-credits.md, read Nov 12, 2026
  9. Perplexity pricingdocs.perplexity.ai/docs/getting-started/pricing, read Nov 12, 2026
  10. SpecStream benchmarksspecstream.ai/benchmarks, read Nov 12, 2026
  11. SpecStream pricingspecstream.ai/pricing, read Nov 12, 2026
  12. SpecStream for agentsspecstream.ai/agents.md, read Nov 12, 2026