# SpecStream API

SpecStream is the live web data engine for teams and agents. This host, api.specstream.ai, serves the Search API, the Extract API, Deep Search and the MCP server. The website is https://specstream.ai.

## Products

- [Search API](https://specstream.ai/docs/search-api): `GET https://api.specstream.ai/v1/web-search`, $5 per 1,000 searches of up to 10 results, and $1 per 1,000 on the fast tier (`speed=fast`).
- [Extract API](https://specstream.ai/docs/extract-api): `POST https://api.specstream.ai/v1/extract`, $0.80 per 1,000 pages. A page with no text is not charged.
- [Deep Search](https://specstream.ai/deep-search): ask in plain words and get every match on the live web as structured rows, each value with its source. Each delivered row is charged 5¢, 15¢ or $1 depending on what the row needs, at a maximum you approve before anything runs.

## Start

1. Sign up at https://specstream.ai/signup: $10 of credit, no card.
2. Make an API key under Agents & API in the workspace. A key that searches or reads pages needs the `finds:spend` scope and a spending limit above 0.
3. Call the API with the key as a Bearer token:

```bash
curl "https://api.specstream.ai/v1/web-search?q=who+won+the+2024+tour+de+france" \
  -H "Authorization: Bearer $SPECSTREAM_API_KEY"
```

How to get a key or an OAuth token, step by step: [https://api.specstream.ai/auth.md](https://api.specstream.ai/auth.md)

## When to use SpecStream

- One fact or a quick answer from the live web: call the Search API.
- The text of pages you already know: call the Extract API with 1 to 20 URLs.
- A complete list or table, with conditions every row has to meet and the same facts on each row: use Deep Search, through the MCP server or the API. It reads the request, names the most it can cost, and runs only after the person approves that maximum.
- A client written for another search API (Serper, Exa, SearXNG, You.com, Parallel and others): change the base address and the key. The addresses are in the agent guide.
- Not the right tool for many people by job title alone: use a contact database for that.

## Reference

- [API docs](https://specstream.ai/docs/api)
- [OpenAPI description](https://api.specstream.ai/openapi.json)
- [Guide for agents](https://specstream.ai/agents.md): every route, reply shape and error, in Markdown
- [MCP server](https://specstream.ai/docs/mcp): `https://api.specstream.ai/mcp` (Streamable HTTP, OAuth or an API key). [Server card](https://specstream.ai/.well-known/mcp/server-card.json)
- [A2A agent card](https://api.specstream.ai/.well-known/agent-card.json)
- [Agent skills](https://specstream.ai/.well-known/agent-skills/index.json)
- [Pricing](https://api.specstream.ai/pricing.md), and the plans at https://specstream.ai/pricing
- Python SDK and CLI: `pip install specstream-client`, then `specstream --help` ([PyPI](https://pypi.org/project/specstream-client/))
- TypeScript SDK: `npm install @specstream/sdk` ([npm](https://www.npmjs.com/package/@specstream/sdk))
- [Changelog](https://specstream.ai/changelog)
- [Everything on the website, for agents](https://specstream.ai/llms.txt)

## Company

- [About SpecStream](https://specstream.ai/about)
- [Security](https://specstream.ai/security)
- [Privacy](https://specstream.ai/privacy)
- [Terms](https://specstream.ai/terms)
- Contact: support@specstream.ai
