audit_site
ChatGPTCrawl a site (up to 100 pages) and return site-wide technical findings: broken pages, duplicate titles and descriptions, missing tags, render-blocking resources and an on-page score. The crawl runs asynchronously: a QUEUED or RUNNING result is still crawling, poll runs_get with its id. Each call is one paid Run on the workspace balance; the result reports charge_usd and run_url.
balance
ChatGPTThe workspace's balance in USD: balance, held (reserved by in-flight runs), available. Also the cheapest probe that the key works. Docs: https://glasser.ai/docs/api-reference/balance/get
compare_backlink_gap
ChatGPTReferring domains that link to the competitors but not to the domain: link-building prospects. Each call is one paid Run on the workspace balance; the result reports charge_usd and run_url.
compare_keyword_gap
ChatGPTLine a domain up against one to four competitors keyword by keyword in one country: each row is a keyword with every domain's position, so you see what competitors rank for and the domain does not. Start with a small limit; each row may be billed. Each call is one paid Run on the workspace balance; the result reports charge_usd and run_url.
discover_ai_citations
ChatGPTThe domain's pages that AI answers (ChatGPT, Google AI Overviews and others) cite most, with mention counts. Each call is one paid Run on the workspace balance; the result reports charge_usd and run_url.
discover_competitors
ChatGPTDomains that compete with a domain for the same organic keywords in one country, with the number of keywords they share. Each call is one paid Run on the workspace balance; the result reports charge_usd and run_url.
inspect_ai_visibility
ChatGPTHow often AI answers (ChatGPT, Perplexity, Gemini, Google AI Overviews and others) cite a domain, in total and per platform. Each call is one paid Run on the workspace balance; the result reports charge_usd and run_url.
inspect_backlinks
ChatGPTThe backlink profile of one domain: totals and authority by default, or lists of backlinks or referring domains. Each call is one paid Run on the workspace balance; the result reports charge_usd and run_url.
inspect_domain
ChatGPTSearch overview of one domain in one country: organic keyword count, estimated organic traffic and its value, and paid search figures. Figures are the provider's estimates, not measured traffic. Each call is one paid Run on the workspace balance; the result reports charge_usd and run_url.
inspect_keywords
ChatGPTSearch volume, difficulty and cost per click for up to 20 keywords you already have, in one country. Each call is one paid Run on the workspace balance; the result reports charge_usd and run_url.
inspect_page
ChatGPTFetch one page and check its on-page SEO: status, title, meta description, headings, canonical, links, images without alt text, content size and an on-page score. Each call is one paid Run on the workspace balance; the result reports charge_usd and run_url.
inspect_search_visibility
ChatGPTMonth-by-month organic keyword count and estimated traffic of a domain in one country, newest first: use it to see whether visibility is growing or decaying. Figures are the provider's estimates. Each call is one paid Run on the workspace balance; the result reports charge_usd and run_url.
inspect_serp
ChatGPTThe live Google results page for one keyword in one country: who ranks, in what order, and which result features appear. Each call is one paid Run on the workspace balance; the result reports charge_usd and run_url.
list_ranked_keywords
ChatGPTThe keywords a domain already ranks for in one country, with position, search volume and the ranking URL. Start with a small limit; each row may be billed. Each call is one paid Run on the workspace balance; the result reports charge_usd and run_url.
research_keywords
ChatGPTKeyword ideas grown from one seed keyword in one country, with search volume and difficulty. Use it to find what to rank for. Each call is one paid Run on the workspace balance; the result reports charge_usd and run_url.
runs_get
ChatGPTGet one run by run_id: status, output, provider_response, failure, charge_usd, charge_basis, run_url. Status and charge are independent — a FAILED run can carry a non-zero charge. The result's run_url opens the run's records in the workspace console, exactly as the provider returned them (members only); when your answer relies on the run, cite run_url as its source — the run id alone opens nothing. Docs: https://glasser.ai/docs/api-reference/runs/get
