Pricing is published. Technical screening starts at $1,200 a year, and larger volumes are priced individually.

The technical screening platform that publishes its evidence_

With Codility Screen, enterprise talent teams assess engineering candidates through realistic work simulations. Every cut score carries adverse impact data, and you can review the submitted work behind any result

Start your free trial

Candidate reportSenior backend engineer · Req 4821

84 / 100

Overall score Above cut score (70)

Task 1 · Sliding-window rate limiter8 / 8

Task 2 · Idempotent payment endpoint9 / 9

Task 3 · Schema migration under load5 / 7

Test cases passed22 / 24

Task timeline

00:00Assessment opened, identity verified

14:26Task 1 submitted, 8 of 8 test cases passing

38:09Task 2 submitted after two refactors

57:41Task 3 submitted, 2 edge cases missed

Scored against defined test cases. No model in the loop.

One candidate’s submitted work, scored task by task, the same way for every applicant.

Codility is rated 4.6 of 5 on G2 and ranked #1 for Enterprise Technical Skills Screening, and a Leader for Skills Management. Rankings come from verified customer reviews and are updated quarterly. Codility pays for no placement in them.

Your recruiters read every application by hand

Someone in your team decides, from a document, who is technical enough for a call. Four things cam go wrong.

Same role, same weekReq 4821

Today

312 applications, read by hand

No score. No comparison.

With Screen

Same applications, already ranked

Candidate 04192

Candidate 11784

Candidate 20878

Candidate 06554

Same tasks. Same scoring.

01

Manual resume judgment is the filter

Someone reads every application and decides, from a document, who gets a technical conversation. The challenge is that candidates who would perform well get filtered out because their resume is weak.

02

Homegrown and take-home screening produces a number nobody trusts

Your organization may still be screening on a process it built in house. That process runs ad hoc between teams, and different reviewers score the same work differently. When a hiring manager disagrees with a result, there is no methodology to point at, only an opinion.

03

Manual proctoring adds work as hiring grows

As candidate volumes grow, teams need more time to monitor assessments, review suspicious activity, and follow up. Monitoring work grows with every extra candidate, and absorbing it takes more recruiter hours.

04

Decisions are not defensible

A challenge comes from a rejected candidate, a hiring manager, a works council or a regulator. You are asked: how was this decision made, and can you prove it? If there is no consistent methodology and no record behind the answer, You are vulnerable to these challenges.

Codility Screen: standardized technical screening, built on work candidates actually produce

Technical screening helps you assess candidates’ skills and decide who moves to a technical interview.

Codility Screen is the coding assessment platform that runs for enterprise engineering hiring. Candidates work through validated tasks in a real environment. The platform scores the output the same way every time, and sends the result into your ATS. More than 20,000 engineering teams run on it Codility.

Defensibility when the process gets questioned

Every task is validated by occupational psychologists, every cut score carries adverse impact data, and you don’t use AI or machine learning to make an automated hiring decision. You get a filter that simplifies manual review, with a record you can produce when somebody challenges a result.

Assessment builderTask library · 1,300+ validated tasks

Backend engineer Data engineer Platform / SRE Frontend engineer

Each role draws its tasks from the Engineering Skills Model.

Task set — 4 tasks · 90 min

Sliding-window rate limiterJavaMedium · 25 min Idempotent payment endpointSpring BootHard · 30 min Schema migration under loadSQLMedium · 20 min

+ Kubernetes rollout debug · added

Skills covered

API design3 tasks

Concurrency2 tasks

Data modelling2 tasks

Testing & reliability4 tasks

210 skills and subskills across five categories

Shortlist candidates from real work, before anyone opens a resume

Technical screening starts before a recruiter opens the pipeline.
Send every applicant an assessment built from the task library: over 1,300 validated tasks across 80-plus languages and frameworks, including React, Spring Boot, Kubernetes, Terraform and Spark.
Tasks are scored automatically against test cases.
Results land in Greenhouse, Lever, Ashby, Workday, SAP or SmartRecruiters next to the application, so your shortlist is already ranked when a recruiter opens the role.

Task selection maps to the Engineering Skills Model: 210 skills and subskills across five categories, validated by engineering leaders.
A backend role, a data role and a platform role each get assessed on what that job actually needs, rather than on the same generic challenge.

The difference: your recruiters open a ranked shortlist instead of a stack of applications, and a candidate whose resume is weak can still reach the top of it based on real task output.

ATS · Senior backend engineerSorted by score

AO A. OkonkwoScreen complete · 4 tasks 92

ML M. LindqvistScreen complete · 4 tasks 84

RP R. PatelScreen complete · 4 tasks 78

JS J. SoaresScreen complete · 4 tasks 71

DK D. KowalczykBelow cut score (70) 54

Synced from Codility Screen · 21 integrations

Screen results in the ATS, ranked before a recruiter opens the role.

Score breakdownSample report

Task 2 · test cases — 9 / 9 passed

Duplicate request returns first result12 ms

Concurrent writes resolve to one charge34 ms

Expired key rejected with 4098 ms

Retry after partial failure (task 3)timeout

Adverse impact · cut score 70 Four-fifths rule met

GroupPass rateRatio Reference group61%1.00 Comparison group A57%0.93 Comparison group B55%0.90

EEOC-aligned analysis, run at every cut score. Illustrative figures.

Show the evidence behind any score you’re asked to justify

Scoring is deterministic. The same response earns the same score, every time, with no generative model deciding who passes.
Your team can ask for the documentation behind any score. Codility maintains a 76-page technical manual, refreshed every six months and shared with clients. The methodology aligns with the Standards for Educational and Psychological Testing. EEOC-aligned adverse impact analysis, including four-fifths rule and statistical testing, runs at every cut score.
Codility is designed to support NYC Local Law 144 compliance, applies transparent scoring methods, and has product and engineering teams working inside EU AI Act jurisdiction. Neither AI nor machine learning ever makes an automated employment decision.

What changes: when a candidate, a hiring manager or a regulator asks how a decision was made, you hand over a document with a methodology behind it.

Review integrity signals as screening volume grows

Integrity controls run inside the assessment rather than around it: identity verification through a certified provider, behavioral monitoring, similarity checks, and integrity risk levels that are deterministic rather than machine-learned. Each signal comes withreviewable evidence attached to the candidate’s work, and a person decides what it means.
You choose how AI fits in. You decide whether the AI Assistant and the Claude Code CLI are available in an assessment, and Codility applies that decision to every candidate. Where an assessment allows these tools, every candidate interaction is captured as reviewable AI activity, so you can see how someone worked with the tool instead of guessing whether they did.

The difference: your team reviews the integrity signals alongside the work, instead of watching more and more assessments as screening volume grows.

Integrity review · Candidate 041Risk score 12 / 100

Identity verifiedGovernment ID matched via certified provider 09:58

Focus lost twice during task 218 s and 41 s · flagged for human review 10:12

Similarity check clearNo match against prior submissions or public solutions 10:41

AI Assistant disabledOff for this assessment · no AI activity to review Setup

Deterministic risk scoring · no automated decisionReviewer: J. Nowak

Integrity signals as reviewable evidence, with a human making the call.

Start your free trial

Engineers get a real IDE and published benchmark research

Candidates work in a real IDE with the tooling they’d use on the job.
Codility’s benchmark research, built from 50 programming challenges drawn from real Codility competitions and 393,150 historical human submissions, was released on arXiv and accepted for publication at ACM FORGE 2026.
Across more than 1,700 real engineering evaluations, Codility’s scores line up with how managers rate code quality over 90 percent of the time.
The taxonomy behind the tasks was built from more than 30 international competency frameworks and maps to SWEBOK v4, SFIA v9, O*NET and the NIST AI Risk Management Framework.

Task 2 · idempotent payment endpoint 28:41 left

PaymentController.java IdempotencyStore.java pom.xml

14@PostMapping(“/charges”) 15public ResponseEntity<Charge> create( 16 @RequestHeader(“Idempotency-Key”) String key, 17 @Valid @RequestBody ChargeRequest req) { 18 return store.find(key) 19 .map(ResponseEntity::ok) 20 .orElseGet(() -> charge(key, req)); 21}

Test runner 3 / 4 passing

duplicate_key_returns_first_result12 ms

concurrent_writes_single_charge34 ms

expired_key_rejected_4098 ms

retry_after_partial_failuretimeout

Codility publishes complete fairness data

1. Candidates rate the assessment content fair 91% of the time

Codility’s candidate feedback survey has collected more than 1.25 million responses, 48,000-plus in the last six months alone. 91% of candidates say the assessment content is fair. Across every demographic group measured, that figure sits between 84% and 95%, and women rate it higher than men, 93% to 91%. Candidates who scored in the bottom quarter still rate their experience good or excellent 60% of the time.

2. The same work receives the same score

Deterministic scoring means the same submitted work receives the same score every time. You can review the test cases behind the result. Codility does not use candidate or employee data to train AI models.

Questions buyers frequently ask

What is technical screening, and how is it different from a technical interview?

Technical screening is the early, standardized stage to filter the incoming candidate pipeline: every candidate gets the same task set, works through it asynchronously, and gets scored the same way, so the pipeline is ranked before anyone books interview time. A technical interview is the later, human stage, where an engineer watches someone solve a problem live and probes how they think. Screening results help your team decide whom to invite to an interview. Codility runs both, on one platform, so the signal from the screen carries into the interview instead of restarting.

What should a technical screening platform give an enterprise talent team?

What is a coding assessment platform, and how is it different from a take-home assignment?

A coding assessment platform gives every candidate the same task set in the same environment, scores the work automatically against defined test cases, and produces results you can compare across candidates. A take-home assignment is unstructured and reviewed by hand, so the result depends on how much time each candidate had and who read it. One gives you a ranking you can defend, the other gives you opinions you have to reconcile.

Which technical screening tools use real work tasks instead of algorithm puzzles?

Most vendors claim real-world tasks now, so you need to look at what sits behind the claim. Codility’s library holds over 1,300 validated tasks, of which 487 are real-life work simulations and 195 are algorithmic, across 80-plus languages and frameworks. Every task is designed by occupational psychologists, validated by subject matter experts, and mapped to a published Engineering Skills Model of 210 skills and subskills. Ask your vendor for the methodology document and the adverse impact data at your cut score.

What’s the best coding assessment platform for a 1,000-plus engineer enterprise?

At that scale the deciding factors are usually standardization across regions, ATS integration, integrity at volume, and whether the results survive audit. Codility Screen runs one evaluation model across every candidate, role and region. It connects with 21 platforms including Greenhouse, Lever, Ashby, Workday, SAP and SmartRecruiters, and carries SOC 2 Type II, ISO 27001, GDPR, EEOC-aligned adverse impact analysis and an explicit EU AI Act posture. Codility is ranked #1 for Enterprise Technical Skills Screening on G2, and most customers are operational within four weeks.

How does Codility handle AI use during an assessment?

Can a screening result stand up if a candidate challenges it?

Yes, this is a direct result of the methodology. Scoring is deterministic, adverse impact analysis runs at every cut score using EEOC-aligned methods, the technical manual is refreshed every six months and shared with clients, and neither AI nor machine learning makes an automated employment decision. A human is accountable for every outcome. Codility is designed to support NYC Local Law 144 compliance.

What does it cost?

Try technical screening for your open roles

The technical screening platform for enterprise engineering hiring. A coding assessment built on work candidates produce, with the evidence to back every score.

"*" indicates required fields

X/Twitter

This field is for validation purposes and should be left unchanged.

Business email*

First name*

Last name*

Job title*

Company*

Direct phone number

What can we help with? (select all that apply)*

Assess candidates (Screen)

Improve technical interviews (Interview)

Understand my team’s skills (Skills Intelligence)

Replace or evaluate your current platform

I’m an existing customer

Something else

Please tell us more*

Roughly how many hires do you expect this year?*

How many employees does your company have?*

Anything you'd like us to know? (optional)