Freehire

Open source alternative to Simplify.jobs

2.4K stars
MIT
Active this week
Last commit today

Last verified Oct 22, 2026

Search 3.3M+ tech job postings indexed directly from company career boards and ATS platforms, deduplicated and filtered by stack, seniority, and location.

Visit website GitHub2.4K Deploy on Hostinger

Read the full breakdownWho it is for, how it works, trade-offs, the quick start and the FAQ

who it's for

Who Freehire is for#

Tech job seekers searching without a paid subscription

Freehire is free to use at freehire.me and searches 3.3M+ deduplicated postings without a paid account. Job seekers who want to filter by tech stack, seniority, and work format can access ATS-sourced data directly with no algorithmic ranking layer pushing sponsored results to the top.

Skip if:

Already using a paid job platform and satisfied with the data quality and search filters.

Developers building on live job market data

The public keyless API and MCP server make freehire a data source for career notification tools, hiring market dashboards, or AI job agents. Developers who need structured, deduplicated job postings can query the full catalogue without negotiating commercial data licenses.

Skip if:

Need historical data older than the current live catalogue, or need coverage beyond tech and IT roles.

Claude Code and MCP host users

The MCP server exposes job search and application tracking as tools callable from Claude Code, Claude Desktop, or any MCP-compatible host. Users who run AI coding assistants can search postings, read facets, and move applications along without leaving their development environment.

Skip if:

Not using Claude Code or another MCP-compatible AI assistant in your regular workflow.

Teams running internal job tracking infrastructure

The self-hosted stack ships with application board, recruiter mail inbox, event ledger, and notification features that a small internal team could deploy on their own servers, replacing the data collection and tracking layers of commercial applicant tracking tools.

Skip if:

Need enterprise ATS features such as multi-user permissions, SLA support, or compliance certifications.

the problem

The problem it solves#

Most job boards operate as aggregators of aggregators: they pull from LinkedIn, Indeed, or third-party feeds and republish the same postings multiple times, sometimes stale by days or weeks. The pain for job seekers is reading the same role four times across different boards, with no clear signal about whether it is still open, where it was originally posted, or who listed it first. Sponsored placements and algorithmic sorting push irrelevant results up, and many premium search filters sit behind monthly subscriptions.

The underlying data loses signal in transit. Most aggregators strip seniority, tech stack, and work format from raw postings and replace them with self-reported values that companies often leave incomplete or wrong. Salary ranges disappear or get averaged away. When a role is posted to Greenhouse, Lever, and LinkedIn simultaneously, there is no canonical record, and the application button may not go where you expect.

how Freehire solves it

How it solves it#

ATS-direct indexing from 320 sources

Every listing is crawled from a company's own ATS page, not from LinkedIn, Indeed, or third-party feeds. Workday, Greenhouse, Lever, Ashby, iCIMS, and 113 other ATS platforms are covered, plus 133 aggregator feeds and 52 direct company career sites, accounting for 3.3M+ unique postings as of October 2026.

Cross-board deduplication into one record

The same role posted to three ATS boards collapses into one entry. Every posting is normalized into a single schema on a stable dedup key, so the job count reflects what you can actually apply to, not how many times the same role has been republished across different platforms.

Faceted search by stack, seniority, and location

Full-text and faceted search over region, work mode, seniority level, skills, and salary. Each facet is derived from curated dictionaries applied to the raw posting text rather than from self-reported values that companies frequently leave blank or wrong.

CLI and MCP server for developer workflows

The CLI installs as a single binary with no Go toolchain required for terminal-based job search and tracking. The MCP server exposes the catalogue and application state to Claude Code and any MCP host: search postings, read facets, and move an application along from within your editor.

Chrome extension with AI application agent

A browser extension puts an AI application agent in Chrome's side panel next to any job posting on any site, not just the ones freehire indexes. It reads the page, scores it against your profile, and can fill the application form. Available from the Chrome Web Store.

Public keyless HTTP API

The job catalogue is accessible over a public, unauthenticated HTTP API. GET https://freehire.me/api/v1/jobs needs no credential. Responses use a consistent shape with data, meta, and error keys. The full endpoint reference and search-filter vocabulary are documented at freehire.me/docs/api.

strengths · trade-offs

Strengths and trade-offs#

Strengths

MIT license with no commercial restrictionsThe entire codebase and crawl pipeline are MIT licensed. You can run your own instance commercially, fork it, or build on top of the API without paying licensing fees or negotiating terms. Unlike API-only commercial job data providers, there is no per-seat or per-query billing on a self-hosted deployment.
Direct source links to company career pagesEvery job listing links to the original posting on the company's own ATS page, not a freehire redirect or an aggregator landing page. When you click through, the application goes straight to the canonical source, which eliminates dead-link problems common on aggregators that cache postings after they close.
Ghost job signal built into the indexFreehire includes a posting-reality signal that flags whether a job is likely still active. This reduces time spent applying to roles that closed before they were removed from public search results. The signal is built into the ingest pipeline, not a manual curation step added on top.
Keyless public API over 3.3M live postingsThe HTTP API at freehire.me/api/v1/jobs is unauthenticated and returns live data. Developers building career apps, job notification tools, or hiring market dashboards can query the full catalogue without signing up for a paid data feed or negotiating a commercial license.

Trade-offs

-Multi-service self-hosting stackRunning your own instance requires Docker Compose, Postgres with pgvector, Meilisearch, Redis, and S3-compatible storage all running simultaneously. This is not a single-binary or one-command deploy. The stack is documented, but it requires server resources and familiarity with container orchestration before the API is reachable.
-AI features need a configured LLM endpointCV tailoring, job fit analysis, enrichment, and the in-app assistant all require an external LLM endpoint configured via environment variables. These features stay disabled when LLM_* env vars are unset, so the base search works without them, but the AI workspace is unavailable until you wire up a provider.
-Relatively new projectThe repository was created in June 2026, making it about four months old at the time of this listing. With 2,413 stars and 6 open issues, early adoption is strong, but production readiness at scale has not been demonstrated over a long run. Teams with strict reliability requirements should evaluate carefully before depending on a self-hosted instance in production.

versus alternatives

Freehire vs alternatives#

Freehire vs Simplify.jobs

Simplify.jobs is a commercial job search and application tracking platform for tech job seekers. Freehire covers the same use case as the open-source, MIT-licensed alternative with a different approach to where job data comes from.

Job data and sourcing

Freehire crawls ATS platforms directly: Workday, Greenhouse, Lever, Ashby, iCIMS, and 113 others. Each listing links to the canonical career page on the company's own domain. Deduplication at ingest means a role posted across multiple boards becomes one record, and the 3.3M+ figure reflects unique open postings after this deduplication.

Commercial aggregators in this category typically pull from existing job boards rather than directly from ATS providers. This introduces an extra aggregation layer between the posting and the user, which can mean stale listings and the same role appearing multiple times under different source labels.

Cost and licensing

Freehire's search is available for free at freehire.me with no account required. The full codebase is MIT licensed and self-hostable via Docker Compose; there are no per-seat fees on a self-hosted instance. Simplify.jobs is a commercial product with free and paid tiers.

Developer tooling

Freehire ships a public, keyless HTTP API over its full catalogue, a CLI binary for terminal-based search, and an MCP server for Claude Code and Claude Desktop. These make it possible to build on top of the job data programmatically without negotiating a commercial data license.

Coverage scope

Freehire focuses on tech and IT roles. Simplify.jobs covers a wider range of industries and roles. For software engineers, data scientists, and related technical roles, freehire's direct ATS coverage is the primary advantage. For roles outside tech, a general-purpose platform with broader industry coverage may be more relevant.

When to choose freehire

Freehire suits technical job seekers who want ATS-direct data with no paid subscription, and developers who need live job data for their own tools via API or MCP. Simplify.jobs may be preferable for non-technical roles, for users who want a polished commercial application management interface, or for teams that need cross-industry coverage.

install · quick start

Quick start#

bash

Install the freehire CLI with the one-line installer; the full self-hosted stack runs via Docker Compose.
```bash
curl -fsSL https://freehire.me/install.sh | sh
```

tech stack · detected from GitHub

What it's built on#

Languages: GoJavaScriptTypeScript
Frameworks: LangChainSvelte
Databases: PostgreSQL
Search: Meilisearch

frequently asked

FAQ#

Is freehire free to use?

The hosted version at freehire.me is free with no account required for search. The self-hosted version is MIT licensed, so there are no per-seat or per-query fees when you run your own instance. Some AI-powered features like CV tailoring and fit analysis require an external LLM endpoint that you configure and pay for separately.

How does freehire differ from LinkedIn or Indeed?

LinkedIn and Indeed aggregate postings from many sources, including recruiter-posted roles, sponsored listings, and third-party feeds. Freehire indexes directly from company ATS platforms, so every listing links to the canonical career page. It also deduplicates the same role posted across multiple boards into one record. The focus is on tech and IT roles rather than the full job market.

Can I run a self-hosted instance of freehire?

Yes. The full stack is available on GitHub under the MIT license. Self-hosting requires Docker Compose, Postgres with pgvector, Meilisearch, Redis, and S3-compatible storage. The make up command in the repository starts the full stack. AI features (CV tailoring, fit analysis, the in-app assistant) are disabled by default and need LLM credentials configured via environment variables to activate.

Does the Chrome extension work on job sites other than freehire?

Yes. The extension runs in Chrome's side panel and works on any page where you are reading a job posting, not just freehire.me. It reads the current page, scores it against your profile, and can fill the application form. It is available from the Chrome Web Store.

What technology does freehire use?

The backend is Go with the Fiber v2 HTTP framework. Storage uses PostgreSQL with pgvector for semantic embeddings. Full-text and faceted search runs on Meilisearch. The frontend is SvelteKit 2 with Svelte 5 runes and Tailwind 4. Redis handles rate limiting. The ingest pipeline uses langchaingo for LLM access with no vendor locked in. The full stack runs locally with Docker Compose.

also worth a look