llms.txt
Compare with other sites' llms.txt files →# Clair
> Clair turns scattered public information into stable developer APIs. One key at https://api.clair.im covers Google Search, Google Maps, Google Shopping, news, jobs, Amazon, Crunchbase, Trustpilot reviews, YouTube transcripts, TikTok and Instagram posts and profiles, and any public page.
Key facts:
- Origin: `https://api.clair.im`. Authenticate with `Authorization: Bearer <key>` from https://clair.im/dashboard/api-keys. One credit-based subscription covers every API. Every path is `/v1/<api>/<resource>`; there is no `engine` parameter.
- Every response starts with `engine`, `family`, and `query` (the request as resolved). Lists add `count`, `total_results` (null when the source shows no total), `results`, and `next_cursor`; send `next_cursor` back as `cursor` for the next page, or pass `page` where the source numbers its pages. `next_cursor` is null on the last page. Single-item endpoints return one named object (`product`, `job`, `article`, …). Ids are named for what they identify (`job_id`, `review_id`, `seller_id`), never a bare `id`. Each call is one billed request.
- Google Search API (`GET /v1/google-search`) returns organic google.com/search hits, 10 per page, up to 10 pages. Ads and `/url?q=` wrappers are stripped.
- Google Maps API: `GET /v1/google-maps/search` scrolls one Maps results list (`q`, `country`, optional `latitude`/`longitude`/`zoom`, no paging) and returns usually 20–70 places with `place_id`, coordinates, rating, `reviews_count`, category, address, `open_state`, phone, website, and booking links; `GET /v1/google-maps/place?place_id=` returns one business with full address, phone, website, `hours[]`, `rating_histogram`, `review_topics`, and top `reviews`. `place.view` is `full` or `limited`; the limited view has no review data.
- Google News API (`GET /v1/google-news`) scrapes news.google.com. Home, search (`q`, optional `when`), topics (`WORLD`, `NATION`, …), geo (`location`), publisher and Full Coverage tokens. Every result includes the publisher URL. Search (`q`) reads up to 10 pages of about 10 results; listings are one page. Article bodies are not hydrated.
- Reuters (`GET /v1/reuters/search`) lists a section (`world` default) or author page — site search is blocked. Bloomberg (`GET /v1/bloomberg/search`) lists a section rail (`latest` default, paged by `cursor`) or site search (`q`, up to 50 pages). Associated Press (`GET /v1/ap/search`) lists a section, topic hub, or site search (up to 20 pages). Each also has an article endpoint (`/v1/reuters/article`, `/v1/bloomberg/article`, `/v1/ap/article`) for authors, dates, tags, images, paragraphs, `access`, and `body_truncated`. One page per request. Not Terminal products.
- Glassdoor (`GET /v1/glassdoor/jobs`) searches public job cards (`workplace_type`, `job_type`, `posted`, up to 10 pages) then `GET /v1/glassdoor/job` reads one listing; `GET /v1/glassdoor/companies` finds an employer's `employer_id` and `slug`, `GET /v1/glassdoor/company` returns its ratings, rating distribution, and newest reviews, and `GET /v1/glassdoor/salaries` returns base and total pay percentiles by job title. Indeed (`GET /v1/indeed/jobs`) searches US job cards (`remote`, `posted`, salary text) then `GET /v1/indeed/job`. Indeed search is one page (about 45 cards), since Indeed shows later pages to signed-in visitors only.
- Amazon searches product cards (`GET /v1/amazon/search`, by keyword or `category_id`, with `sort_by`, `min_price`, `max_price`, `brand`, `prime`, `min_rating`, `condition`, and `deals`; cards flag `sponsored`, `is_prime`, `coupon`, and `other_offers`), reads one ASIN (`GET /v1/amazon/product`), its top reviews (`/v1/amazon/reviews`), every seller offer (`/v1/amazon/offers`), maps a UPC/EAN/ISBN to an ASIN (`/v1/amazon/lookup`), reads seller profiles and storefronts (`/v1/amazon/seller`, `/v1/amazon/seller-products`), best sellers (`/v1/amazon/bestsellers`), and the category tree (`/v1/amazon/categories`). Every Amazon endpoint takes `country` (us, gb, ca, au, in, de, es, nl, be, jp) and `zip_code` for the delivery location. Crunchbase reads one organization at `GET /v1/crunchbase/organization` or one person at `GET /v1/crunchbase/person`. Fields Crunchbase hides stay in `locked_fields`.
- Web Scraper (`GET /v1/scrape?url=`) returns any public http(s) page as `page.html` with `page.final_url`, the page's own `page.status`, `page.title`, and `page.rendered`. A direct fetch goes first, then a browser that gets past bot checks; `render=true` goes straight to the browser for JavaScript-built pages. It does not parse the page.
- Missing source values stay `null`. Only the Web Scraper takes an arbitrary URL; every other API takes documented query parameters.
- Suitable uses: retrieval, monitoring, research, recruiting products, reputation workflows. Not a full-article redistribution archive or a private vendor API.
## Documentation
- [Documentation overview](https://clair.im/docs)
- [Quickstart](https://clair.im/docs/quickstart)
- [Authentication, credits, and errors](https://clair.im/docs/authentication)
- [Web Scraper API](https://clair.im/docs/scrape-api/web-scraper)
- [Amazon API](https://clair.im/docs/products-api/amazon)
- [Google Search API](https://clair.im/docs/search-api/google-search)
- [Google Maps API](https://clair.im/docs/search-api/google-maps)
- [YouTube API](https://clair.im/docs/video-api/youtube)
- [TikTok API](https://clair.im/docs/video-api/tiktok)
- [Instagram API](https://clair.im/docs/video-api/instagram)
- [Glassdoor API](https://clair.im/docs/jobs-api/glassdoor)
- [Indeed API](https://clair.im/docs/jobs-api/indeed)
- [Google News API](https://clair.im/docs/news-api/google-news)
- [Google Shopping API](https://clair.im/docs/products-api/google-shopping)
- [Trustpilot API](https://clair.im/docs/reviews-api/trustpilot)
- [Crunchbase API](https://clair.im/docs/companies-api/crunchbase)
- [Reuters API](https://clair.im/docs/news-api/reuters)
- [Bloomberg API](https://clair.im/docs/news-api/bloomberg)
- [Associated Press API](https://clair.im/docs/news-api/ap)
## OpenAPI specs
OpenAPI 3.1, one spec per API. All 16 APIs are also in one [Postman collection](https://www.postman.com/clair-api-team/clair-api/collection/58797287-3f95b24e-229a-479c-b33d-58f111c46392).
- [Web Scraper API](https://clair.im/openapi/scrape.json)
- [Amazon API](https://clair.im/openapi/amazon.json)
- [Google Search API](https://clair.im/openapi/google-search.json)
- [Google Maps API](https://clair.im/openapi/google-maps.json)
- [YouTube API](https://clair.im/openapi/youtube.json)
- [TikTok API](https://clair.im/openapi/tiktok.json)
- [Instagram API](https://clair.im/openapi/instagram.json)
- [Glassdoor API](https://clair.im/openapi/glassdoor.json)
- [Indeed API](https://clair.im/openapi/indeed.json)
- [Google News API](https://clair.im/openapi/google-news.json)
- [Google Shopping API](https://clair.im/openapi/google-shopping.json)
- [Trustpilot API](https://clair.im/openapi/trustpilot.json)
- [Crunchbase API](https://clair.im/openapi/crunchbase.json)
- [Reuters API](https://clair.im/openapi/reuters.json)
- [Bloomberg API](https://clair.im/openapi/bloomberg.json)
- [Associated Press API](https://clair.im/openapi/ap.json)
## Products and resources
- [API catalog](https://clair.im/apis)
- [Use cases](https://clair.im/use-cases)
- [SERP and search](https://clair.im/use-cases/serp-search)
- [LLM grounding](https://clair.im/use-cases/llm-grounding)
- [Media monitoring](https://clair.im/use-cases/media-monitoring)
- [Job market analytics](https://clair.im/use-cases/job-market-analytics)
- [Review intelligence](https://clair.im/use-cases/review-intelligence)
- [Comparisons](https://clair.im/vs)
- [Glossary](https://clair.im/glossary)
- [Resource catalog](https://clair.im/resources)
- [Pricing](https://clair.im/pricing)
## Access and policies
- [Terms of service](https://clair.im/legal/terms)
- [Privacy policy](https://clair.im/legal/privacy)
- Contact: contact@serinlabs.com
