ClawJobsPublic API

Job boards have historically been walled gardens. LinkedIn requires enterprise partnerships for API access. Indeed deprecated its Job Search API entirely. If you wanted to build tools around job data, your options were expensive aggregators or brittle scraping.

The ClawJobs API changes that for AI and crypto jobs. We pull listings directly from company ATS platforms - Greenhouse, Lever, Ashby, Workable - and expose them through a clean REST API. No partnership applications. No enterprise contracts. Sign in, grab an API key, and start building.

The API currently tracks positions at 500+ companies including OpenAI, Anthropic, Google DeepMind, Coinbase, Uniswap, and hundreds more across the AI and crypto ecosystem.

Getting started takes under a minute:

1. Sign in with Google to create your account
2. Generate an API key from your profile page
3. Make your first request:

curl "https://api.clawjobs.cc/v1/jobs?category=ai&limit=5" \
  -H "Authorization: Bearer clawjobs_your_key"

See the full API reference for all endpoints, parameters, and response formats.

Seven endpoints cover the core data:

The free tier works without any API key - just start making requests. It returns basic fields so you can explore the data and build prototypes. Pro unlocks the full dataset.

AI-Powered Job Search Agent

The most compelling use case right now. Build an agent that monitors ClawJobs for new positions matching your skills, scores them against your resume, and either alerts you or applies automatically. The search endpoint supports keyword queries, and the Pro tier gives you full job descriptions for semantic matching with embeddings or LLMs.

A basic flow looks like this:

1. Poll GET /v1/search?q=machine+learning&remote=true on a schedule
2. Compare new results against previously seen job IDs
3. For new jobs, fetch full details with GET /v1/jobs/:id
4. Use an LLM to score the job description against your resume
5. If the score passes your threshold, extract the apply URL and submit

Tools like Sonara and JobCopilot charge $20-80/month for this workflow on generic job boards. With the ClawJobs API, you can build a version focused specifically on AI and crypto roles - the sectors where job descriptions are most technical and where generic auto-apply tools produce the worst matches.

Hiring Dashboard for Recruiters

Recruiting teams tracking AI or crypto talent can use the API to build internal dashboards. The /v1/jobs/stats endpoint gives you real-time market snapshots - how many remote ML engineer roles are open today, which cities have the most demand, what departments are growing fastest. Track competitor hiring by monitoring specific companies with the /v1/companies/:id endpoint.

Job Board Integrations

If you run a niche community, newsletter, or Slack workspace, you can pull relevant jobs from ClawJobs and display them alongside your content. Filter by category (ai, defi, infrastructure), by department, or by location. The data updates daily as we scrape fresh listings from ATS platforms.

Market Research and Talent Analytics

VCs evaluating portfolio companies, analysts tracking industry trends, and job seekers assessing market conditions can all benefit from structured job data. Track which companies are scaling specific teams, when new categories of roles emerge (like "AI safety researcher" or "LLM fine-tuning engineer"), and how remote vs. on-site ratios shift over time.

Slack and Discord Bots

Build a bot that posts new matching jobs to a channel. Set up a cron job that queries the API every hour, diffs against previously seen listings, and pushes new matches to your team's Slack or Discord. Especially useful for developer communities where members want job notifications without leaving their workflow.

The job data API space has consolidated significantly. The two largest job boards - LinkedIn and Indeed - have both restricted or deprecated public API access. Here is how the landscape looks in 2026:

The auto-apply market is growing fast. Tools like Sonara, LazyApply, and JobCopilot charge $20-80/month for automated job applications. But the biggest complaint from users is poor matching quality - generic tools send out irrelevant applications because they lack domain-specific understanding.

By building on the ClawJobs API, you can create a focused agent that understands the nuances of AI and crypto hiring. Here is a Python example using the search endpoint with an LLM for scoring:

import requests
import anthropic

API_KEY = "clawjobs_your_key"
BASE = "https://api.clawjobs.cc/v1"
headers = {"Authorization": f"Bearer {API_KEY}"}

# Search for matching jobs
resp = requests.get(
    f"{BASE}/search",
    params={"q": "ml engineer", "remote": "true", "limit": 20},
    headers=headers,
)
jobs = resp.json()["data"]

# Score each job against your resume
client = anthropic.Anthropic()
for job in jobs:
    detail = requests.get(
        f"{BASE}/jobs/{job['id']}", headers=headers
    ).json()["data"]

    msg = client.messages.create(
        model="claude-sonnet-4-20250514",
        max_tokens=200,
        messages=[{
            "role": "user",
            "content": f"""Score this job 0-100 for fit.
Resume: [your resume text]
Job: {detail['title']} at {detail['company_name']}
Description: {detail.get('description', 'N/A')}
Reply with just the score and one sentence."""
        }],
    )
    print(f"{detail['title']} - {msg.content[0].text}")

This is a starting point. A production agent would store seen job IDs to avoid re-processing, run on a cron schedule, and integrate with an application submission step using the url field from the Pro tier response.

The API is built with Rust and Axum, backed by PostgreSQL with OpenSearch for full-text queries. All responses use a consistent JSON envelope:

{
  "data": [...],
  "meta": {
    "total": 2847,
    "limit": 20,
    "offset": 0,
    "has_more": true
  }
}

Authentication supports two header formats:

• Authorization: Bearer clawjobs_your_key
• X-API-Key: clawjobs_your_key

Rate limits use a rolling 60-second window tracked per API key via Redis. When you exceed your limit, the API returns HTTP 429 with a standard error body. Errors follow a consistent format across all endpoints:

{
  "error": {
    "code": "RATE_LIMITED",
    "message": "Too many requests"
  }
}