Gem

Gem is recruiting software with several AI agents. Two of them find and rank people. The AI Sourcing Agent finds people who have not applied. AI Application Review ranks the people who have. Both are paid add-ons, and Gem quotes the price by company size. In Application Review, a recruiter makes every advance or reject decision.

Gem finds people and ranks applicants

In one line

Gem’s AI Sourcing Agent searches more than 800 million public profiles and your own past candidates. AI Application Review scores each applicant against your criteria. A recruiter decides who moves forward.

Buy it if

You want one tool to find people and to rank applicants. You run Greenhouse, Workday or iCIMS, or you will use Gem’s own tracking system. You accept Terms of Service with a permanent licence to your candidate data.

Do not buy it if

You need a published price for the AI features. You need a promise that you can get all your data back when you leave.

Price

Gem quotes by your headcount. The AI features come with an AI-Powered seat, bought through an account manager. Companies under 30 employees get Gem All-in-One free for six months.

What Gem’s two AI tools do for a recruiter

A recruiter types the person they want in ordinary words. Gem turns that into a search across more than 800 million public profiles, your past candidates and your current pipeline. A browser add-on saves profiles from LinkedIn and more than 20 other sites. Each result gets a match score and a reason.

Gem adds a verified email address to each profile. Its email delivery rate is 98%. Gem drafts messages for each person and can send them from the recruiter’s own email inbox. A recruiter reviews and approves the drafted messages before they go out. After approval, steps set to Automatic send by themselves. A reply stops the sequence.

AI Application Review works on people who applied. It scores each applicant against the criteria you set. It reads the CV, past contact with the person and your candidate database. Each applicant gets a plain-language summary of why they fit or do not. The overall score is an average of the score for each criterion. You switch ranking on role by role.

The recruiter can advance, reject or keep applicants for later in bulk. Before scoring, Gem removes the applicant’s name, email and LinkedIn address from the CV only. The AI Fraud Detection Agent runs in the background and flags suspicious applicants. It is a separate purchase.

Application Review needs three things in place. It needs Gem’s candidate database, the Application Review product and one tracking system. That system is Gem’s own, Greenhouse, Workday, SuccessFactors or iCIMS. Searching past applicants works only on Gem’s own system or on certain other systems. Ask Gem whether yours is one of them.

What Gem and its customers have seen

Gem

Recruiters act on 1 in 2 of the candidates the AI Sourcing Agent surfaces.

Gem, high-volume roles

Customers hiring for high-volume roles see review times 60% faster.

Faire, 1,000 to 5,000 employees

A two and a half month trial of AI sourcing turned up more than 7,700 new candidates. They were spread across 8 open roles.

Workiva, 2,800 employees

Application Review ran on thousands of applications. The time to close each one fell by 18 days.

Mission Cloud, 300 employees

Two recruiters made 43 hires in 90 days using Gem’s AI ranking and other AI tools.

Tropic

Application Review cut screening from 3 to 4 hours down to about 30 minutes.

Gem has raised $148 million in total. More than 1,200 hiring teams run their recruiting on Gem. A $100 million Series C in September 2021 valued the company at $1.2 billion.

Who Gem suits

You want to find people who have not applied

AI Sourcing has to be switched on for your account. If it is not, Gem’s customer success team handles it. Ask for it in the quote.

You run Greenhouse, Workday or iCIMS

Application Review sits on top of Greenhouse, Workday or iCIMS. Applicants flow into Gem and get ranked. The scores are saved back into that system. SuccessFactors and Gem’s own tracking system work with it too.

Your recruiters want to see why a person scored well

Every score comes with a plain-language reason. Your team can read the reason and filter applicants by experience, location or source.

You hire in languages other than English

Gem tests and develops its AI features in English only. Ask Gem what happens to a CV or a search in another language before you plan around it.

Where Gem stops, from your data to your contract

The Terms of Service give Gem a permanent licence to your candidate data

Gem’s Terms of Service give Gem a licence to Candidate Data that never ends and cannot be taken back. Gem may use, copy, change and share it. Candidate Data means any customer data about a person you consider for a job.

Where your data goes when the AI runs

Gem’s AI runs on Microsoft Azure OpenAI. Gem does not use customer data to train the AI. Gem removes names, locations and LinkedIn handles from profile data before it goes to the model. Microsoft may keep inputs and outputs for up to 30 days to detect abuse. Microsoft does not use them for training. Gem uses data centres in the United States only.

Gem can change its prices at any time

Under the Terms of Service, Gem may change its price list and add new fees at any time. If you cancel after a price change, the fees for your current payment term are not refunded. Subscriptions renew automatically. Under the Enterprise agreement, each renewal runs one year unless one side gives 30 days’ notice.

You carry the legal risk of how you use the AI

Under the Enterprise agreement, it is your responsibility to know hiring law, including the law on illegal discrimination. You are solely responsible for how you use the AI’s output and whether that use is legal. You must also defend Gem against outside claims that your use broke a law. Each side’s liability is capped at what you paid in the previous 12 months. The cap does not cover indemnification, so your duty to defend Gem is not capped.

Leaving may cost you the data held in the AI features

Under the Enterprise agreement, Gem has no duty to keep your data or give it back after the term. The standard retention period is 90 days after your last contract ends. Data you sent to the AI features may not be retrievable or exportable when the AI Feature Terms end. Ending the AI Feature Terms brings no refund, credit or other payment. Exports come as spreadsheet files (CSV). During the term you can ask for deletion or export of candidate data. You can also set your own retention period in the admin screen.

What the contract asks of you

Under the Terms of Service, you must take part in Gem’s press announcements and case studies. Under the Enterprise agreement, Gem may name you as a customer in its advertising and marketing. Gem will not sign your own data processing agreement. It offers its standard one. Disputes under the Terms of Service go to individual arbitration, and class actions are barred.

Gem builds its database of people from public sources

Gem builds its profile database from public sources and from information its customers upload. The sources include Indeed, GitHub, Facebook, Twitter and SeekOut. Gem relies on legitimate interests, a lawful basis in European data law, to process it. Gem may also sell or share personal information with customers that want to consider a person as a candidate. A person can opt out, or ask for erasure in some cases. A person anonymised by your retention rules can be sourced again by Gem users.

The bias audit tested the default weights only

New York City’s Local Law 144 covers software whose score largely shapes or replaces a hiring decision. Employers in the city must have that software audited for bias every year. BABL AI, an outside firm, audited Gem’s AI ranking as if the law applied. It did not decide that the law does apply. Gem does not consider its tools to be covered by the law. The audit did not certify the model as free of bias. Gem carried out the testing itself, on data to September 2024. It tested at the default weights for each criterion. A recruiter can star a criterion to give it ten times the weight.

Some applicants never get a score

The AI does not score an application with no CV attached. The AI can make mistakes. It is meant to help review inbound applications, not replace the review.

AI sourcing searches have size limits

Each AI sourcing search returns at most 1,000 candidates. A search covering more than 20,000 candidates may return less relevant results.

Some steps in Gem run without a recruiter

In Gem’s own tracking system, knockout questions can reject a candidate and send the rejection email. Saved search automations add people to message sequences. Recruiters are not notified when they do. With Premium Email Finding, a sequence sends once an email is found.

What Gem costs

Gem quotes by your number of employees. It has two plans. Gem + Your ATS is the plan for teams that keep their own applicant tracking system (ATS). The other plan is All-in-One. On Gem + Your ATS, scheduling, the AI Fraud Detection Agent and analytics are add-ons.

Standard seats carry the base features. The AI tools come with an AI-Powered seat. AI Sourcing and Application Review are paid add-ons, bought through your account manager. Ask Gem what the AI-Powered seat adds to your quote.

A company under 30 employees gets All-in-One free for six months, then 50% off a paid plan. A company of 30 to 100 employees gets reduced rates from day one, with no sales call. Gem’s published per-user prices are for recruiting firms only. Company talent teams are not eligible for those plans.

Retention rules, including periods by country, need the Advanced CRM package. CRM is Gem’s candidate database. Ask Gem whether your plan includes Advanced CRM.

You could also find and screen people with agents built on your own systems. We have mapped what that looks like. The HR Agent Map

Nine questions to put to Gem

  1. Can you narrow the Terms of Service licence to running our account? Will you write that into our order form? Good answer: A written list of permitted uses, with no licence to candidate data beyond running our account.
  2. The audit tested default weights. Do you have results for the weights we plan to use, including starred criteria? Good answer: A test at our weights, run by the outside auditor, and a plan to repeat it each year.
  3. Will you cap price changes during our term? Will you refund the unused months if we cancel after a change? Good answer: A fixed price for the term, in writing, and a refund of unused months if prices rise.
  4. Does Application Review work with Lever? Do scores go back into SuccessFactors as well? Good answer: One written list that names each system and says whether scores are saved back.
  5. What price will you quote at our headcount for the AI-Powered seat, Application Review and email finding? Good answer: A written quote that lists each add-on, each seat type, the email finding allowance and the renewal price.
  6. When our contract ends, how do we get back the scores and summaries from the AI features? Good answer: A named file format, a time window, and a promise to delete the rest after export.
  7. Your data exports include fields for predicted gender and predicted race or ethnicity. How are they produced? Can we switch them off? Good answer: A clear method, a setting to switch them off, and no use of them in scoring.
  8. How do you answer a person who asks why they appear in your database and how to be removed? Good answer: A named process, a response time, and removal that covers every Gem customer account.
  9. Which contract will we sign, the Terms of Service or the Enterprise agreement? Which terms win if they differ? Good answer: A written statement of which terms govern.

Alternatives to Gem

The alternatives are LinkedIn Recruiter, SeekOut, hireEZ, Findem, Pin, Greenhouse and Workable, or buying nothing.

LinkedIn Recruiter is the paid LinkedIn tool for finding people who have not applied. Its AI add-on, Hiring Assistant, is sold on top of a Recruiter contract. LinkedIn is rolling it out gradually, so a customer may not have it yet. SeekOut searches nearly one billion public profiles. Its Applicant Evaluation and AI Screening reviews applicants against the role requirements.

hireEZ runs on top of your tracking system. Every candidate gets a screening conversation by voice or chat using your questions, and you decide who moves forward. Findem draws on more than 850 million profiles. It connects to your tracking system and candidate database to keep candidate data current. It does not replace them.

Pin works alongside your tracking system. Its AI recruiting agent covers sourcing, screening, messages and booking. Staffing agencies, boutique firms and independent recruiters use Pin. Pin also offers a version for in-house recruiting teams.

Greenhouse sells Real Talent. Its filters sort applications by how well they match the criteria the employer sets. It never advances or rejects a candidate. It works much like Application Review. Real Talent also targets fraud and spam. Greenhouse sells Voice AI interviews in 11 languages. Voice AI works with or without the Greenhouse system.

Workable Agent is built into Workable’s tracking system. It runs the first steps of hiring, from sourcing and screening to a shortlist. It can be added to any Workable plan and is paid for with prepaid AI credits.

Buying nothing means no licence over your candidate data and no renewal to track. It also means searching and sorting stay manual. Our guide to AI sourcing and screening covers what the rest of the market offers.

Sources Everything about Gem comes from these addresses. So does everything about the alternatives. Each line gives the date the address was read.

https://www.gem.com/product/ai-sourcing · Gem AI Sourcing · read 8 Oct 2026
https://www.gem.com/product/ai-app-review · Gem AI Application Review · read 8 Oct 2026
https://www.gem.com/product/ats · Gem applicant tracking system · read 8 Oct 2026
https://www.gem.com/ai-recruiting-agents · Gem AI recruiting agents · read 8 Oct 2026
https://www.gem.com/compliance/security · Gem security page · read 8 Oct 2026
https://www.gem.com/pricing · Gem pricing · read 8 and 9 Oct 2026
https://www.gem.com/staffing/pricing · Gem pricing for recruiting firms · read 8 and 9 Oct 2026
https://www.gem.com/company · Gem company page · read 8 Oct 2026
https://techcrunch.com/2021/09/28/recruiting-platform-gem-gains-unicorn-status-with-100m-raise-to-change-the-way-companies-hire · TechCrunch, Gem funding, 28 September 2021 · read 8 Oct 2026
https://www.gem.com/enterprise-msa · Gem Enterprise master services agreement · read 8 and 9 Oct 2026
https://www.gem.com/compliance/terms · Gem Terms of Service · read 8 and 9 Oct 2026
https://www.gem.com/compliance/ai-feature-terms · Gem AI Feature Terms · read 8 and 9 Oct 2026
https://www.gem.com/compliance/privacy · Gem privacy policy · read 8 and 9 Oct 2026
https://www.gem.com/compliance/privacy-policy-non-b2b-customers · Gem privacy policy for people who are not customers · read 8 and 9 Oct 2026
https://help.gem.com/external/gem-ai-legal-privacy-compliance-faqs · Gem AI legal, privacy and compliance questions · read 8 and 9 Oct 2026
https://help.gem.com/external/gem-ai-features-overview · Gem AI features overview · read 8 Oct 2026
https://help.gem.com/external/ai-powered-app-review · Gem help, AI-powered Application Review · read 8 Oct 2026
https://help.gem.com/external/getting-started-with-ai-sourcing · Gem help, getting started with AI sourcing · read 9 Oct 2026
https://help.gem.com/external/gem-ai-rediscovery-overview-getting-started · Gem help, AI rediscovery · read 8 Oct 2026
https://help.gem.com/external/sequences-overview-external · Gem help, sequences · read 8 Oct 2026
https://help.gem.com/external/premium-email-finding-automatically-send-sequence-when-email-is-returned · Gem help, Premium Email Finding · read 8 Oct 2026
https://help.gem.com/external/prospect-search-saved-search-automations · Gem help, saved search automations · read 8 Oct 2026
https://help.gem.com/databases/gem-help-center/gem-ats-knockout-questions · Gem help, applicant tracking knockout questions · read 8 and 9 Oct 2026
https://help.gem.com/external/data-retention-policy · Gem help, data retention · read 8 and 9 Oct 2026
https://help.gem.com/external/gem-data-export-formatschema · Gem help, data export format · read 8 and 9 Oct 2026
https://help.gem.com/external/gdpr · Gem help, GDPR · read 8 Oct 2026
https://www.gem.com/case-study/Faire · Gem case study, Faire · read 8 Oct 2026
https://www.gem.com/case-study/workiva · Gem case study, Workiva · read 8 Oct 2026
https://www.gem.com/case-study/mission-cloud · Gem case study, Mission Cloud · read 8 Oct 2026
https://www.gem.com/case-study/tropic · Gem case study, Tropic · read 8 Oct 2026
https://talent.okta.com/rs/987-NDK-658/images/gem-independent-bias-audit.pdf · BABL AI, summary of the bias audit of Gem AI Ranking · read 8 and 9 Oct 2026
https://www.linkedin.com/help/recruiter/answer/a7104929 · LinkedIn Help, Hiring Assistant · read 8 Oct 2026
https://business.linkedin.com/talent-solutions/recruiter · LinkedIn Recruiter · read 8 Oct 2026
https://www.seekout.com/solutions/recruit/ · SeekOut Sourcing · read 8 Oct 2026
https://www.seekout.com/solutions/sam/ · SeekOut Applicant Evaluation and AI Screening · read 8 Oct 2026
https://hireez.com/ez-agent/ · hireEZ EZ Agent · read 8 Oct 2026
https://www.findem.ai/products/talent-sourcing · Findem Talent Sourcing · read 8 Oct 2026
https://www.pin.com/features/ai-recruiting-agent/ · Pin AI recruiting agent · read 8 Oct 2026
https://www.pin.com/pricing/ · Pin pricing · read 8 Oct 2026
https://www.pin.com/built-for/recruiting-teams/ · Pin for in-house recruiting teams · read 8 Oct 2026
https://www.greenhouse.com/real-talent-candidate-matching · Greenhouse Real Talent · read 9 Oct 2026
https://www.greenhouse.com/voice-ai-interviewing · Greenhouse Voice AI · read 9 Oct 2026
https://investor.wedbush.com/wedbush/article/gnwcq-2026-3-13-workable-launches-agentic-ai-for-hiring-with-new-workable-agent · Workable launch release for Workable Agent · read 9 Oct 2026
https://www.workable.com/workable-ai · Workable AI · read 9 Oct 2026
https://www.workable.com/pricing · Workable pricing · read 9 Oct 2026
https://legistar.council.nyc.gov/View.ashx?M=F&ID=10369162&GUID=F25C6911-8267-48AF-9048-180D44091624 · New York City Local Law 144 · read 9 Oct 2026