Jobs API in Python: pip install to live postings in 5 minutes

The official jobspipe package is on PyPI. This is the whole quickstart: install, export a key, eight lines to live postings.

pip install jobspipe
export JOBSPIPE_API_KEY=jp_live_...   # free at jobspipe.dev/signup

from jobspipe import Jobspipe

client = Jobspipe()  # reads JOBSPIPE_API_KEY

results = client.jobs.search(
    description_or=["python", "django"],
    job_country_code_or=["US"],
    remote=True,
    posted_at_max_age_days=7,
    limit=25,
    include_total_results=True,
)

print(f"{results.metadata.total_results} matches")
for job in results.data:
    print(f"{job.job_title} @ {job.company}")

The filters that matter

• job_title_or / job_title_not - match or exclude on title; description_or searches body text for skills.
• posted_at_max_age_days - recency window. Combined with a daily run this is a new-jobs feed in four lines.
• remote, job_country_code_or, job_seniority_or - the workforce slicers.
• company_name_or - watch specific accounts, the core of hiring-signal monitoring.

Every response row carries parsed salary ranges where present, ISCO-08 occupation codes, and company enrichment - the full shape is in the search API guide.

Size before you spend

Billing is one credit per job returned - an empty result costs nothing. Before pulling a big result set, pass blur_company_data=True: the response returns masked employer fields but real titles, locations, salaries, and the full match count, at zero credit cost. Check results.metadata.total_results, tune the filters, then run the unblurred query.

Paginate with the cursor

Use results.metadata.next_cursor and pass it back as cursor rather than using offsets - keyset pagination stays stable while new postings land, so you never pay twice for a row that shifted pages. If your consumer is an agent rather than a script, the same corpus is queryable over MCP - see the MCP server post.