Lever.co Jobs Scraper & API — Any Company Board
Pricing
from $2.00 / 1,000 job results
Scrape any company's Lever job board (jobs.lever.co) on demand — slugs, board URLs or careers-page URLs in, clean JSON out: title, salary, location, remote, apply URL. Delta mode bills only new postings; optional BYOK AI enrichment. No login, no proxies, no HTML parsing.
Company slugs as used on their Lever board — full jobs.lever.co URLs and plain company careers-page URLs (e.g. acme.com/careers) are also accepted; the Lever board is auto-detected from the page. Leave empty to run a built-in default list of well-known Lever boards (palantir, spotify, aircall, qonto, contentsquare).
Optional case-insensitive substring match on the job title (e.g. engineer).
Drop postings whose title contains any of these substrings (case-insensitive). Example: [intern, staffing].
Optional case-insensitive substring match on the location (e.g. remote, london).
Only keep postings first published within this many days. Leave empty (or 0) for no date filter. When this is set, postings the source didn't give a posting date for are dropped rather than guessed at.
Keep only postings Lever flags workplaceType=remote. Postings marked on-site/hybrid, or where Lever didn't state a workplace type, are excluded.
Include a plain-text description snippet per posting. Turn off for lighter, faster output.
Attach the untouched Lever posting payload as a raw field on every record — for power users who want fields this Actor doesn't map. Off by default (adds significant size to each row).
How much of each posting to output. full = every field (default). compact = drops the heavy free-text fields (snippet, raw). minimal = just the essentials (title, company, location, remote, salary, url, postedAt, globalId). Opt-in AI / delta fields are always kept.
Cap postings returned per company. Each posting returned is a billed result, so this keeps one large board from dominating the run's cost. 0 = no cap.
Hard cap on total postings returned across all companies. Each posting returned is billed as a paid result — raise this only as high as you actually need. 0 = no cap (not recommended for large company lists).
Delta/monitoring mode: only output postings not seen in a previous run that also had this flag on. Already-seen postings are dropped before push (not billed), so this is the cheapest way to run this Actor on a cron schedule and only pay for genuinely new postings. State is tracked per Actor in a dedicated key-value store, keyed by each posting's globalId. See README “Delta mode / monitoring”.
Adds aiKeySkills / aiExperienceLevel / aiWorkArrangement / aiVisaSponsorship to every posting via the Anthropic or Mistral API (pick which below). Requires your own API key for the provider you pick — billed separately by that provider, not by this Actor (see README “AI enrichment”). If turned on without a matching key, enrichment is skipped (with a warning) and postings are still returned normally, without the ai* fields.
Add titleEnglish / snippetEnglish to every posting, machine-translated via the AI provider you pick below (BYOK). Useful for non-English Lever boards (French, German, etc.); already-English postings pass through unchanged. Requires a matching API key — skipped (with a warning) if none is provided.
Add company-level aiCompanyIndustries / aiCompanySummary to every posting, resolved once per company via the AI provider (BYOK). Grounded in the company's own postings; companies the model doesn't recognise get empty values rather than a guess. Requires a matching API key — skipped (with a warning) if none is provided.
Which AI provider runs enrichment (only relevant when “AI enrichment” is on). anthropic (default) uses Claude via anthropicApiKey. mistral uses a Mistral model via mistralApiKey instead — pick this if you'd rather bring a Mistral key than an Anthropic one. openai uses a GPT model via openaiApiKey.
Your Anthropic API key (sk-ant-…). Only used when “AI enrichment” is on and aiProvider is anthropic; billed separately by Anthropic. Not required unless aiEnrichment is on.
Claude model used for AI enrichment when aiProvider is anthropic (only relevant when “AI enrichment” is on). Haiku is fast and inexpensive; Sonnet gives higher-quality extraction on longer, more nuanced descriptions.
Your Mistral API key. Only used when “AI enrichment” is on and aiProvider is mistral; billed separately by Mistral. Not required unless aiEnrichment is on with aiProvider=mistral.
Mistral model used for AI enrichment when aiProvider is mistral (only relevant when “AI enrichment” is on). Small is the default — testing found it matches larger Mistral models on this extraction task (small, well-scoped fields) at a fraction of the cost.
Your OpenAI API key (sk-…). Only used when “AI enrichment” is on and aiProvider is openai; billed separately by OpenAI. Not required unless aiEnrichment is on with aiProvider=openai.
OpenAI model used for AI enrichment when aiProvider is openai (only relevant when “AI enrichment” is on). Defaults to gpt-4.1-mini — cheap, fast and ample for this extraction task. Any chat-completions model works, including the gpt-5 family (e.g. gpt-5.4-mini).
![Lever Jobs [$0.99] Scraper & API — Any Company avatar](https://images.apifyusercontent.com/_O59p-0aJZTFwFCTm0uCphWnqZ64TLsp4-QoJcjUXFQ/rs:fill:76:76/cb:1/aHR0cHM6Ly9hcGlmeS1pbWFnZS11cGxvYWRzLXByb2QuczMudXMtZWFzdC0xLmFtYXpvbmF3cy5jb20vbmduOHZSRVNUa3NleVdFNnAtYWN0b3ItTU9EU2JIY0JqbUxKRnRqRnYtb1ZqU1BIZmRaaC1sZXZlci5wbmc.webp){kind=avatar}
