AI skills assessments compared
Several hiring tools now test AI skills. They differ most in one thing: whether the candidate actually works with an AI on a task, or answers questions about how they would.
At a glance
| What the candidate does | Price | |
|---|---|---|
| SHOWCASE | Does a real task with Claude in a private workspace, with files that have problems planted in them. 20 minutes, or 45 for engineering. Every message, command, and file is recorded. | $5 per test. No subscription. |
| Bryq | A simulation-based task, about 15 minutes, “tool-agnostic.” Scored 0–100 on five dimensions. | From $69 a month, billed annually ($828 a year), for companies of 1–15 employees. |
| TestGorilla | A library of AI fluency tests, mostly multiple choice and about 10 minutes, plus AI-led interviews and scenario tasks. | From $215 a month, billed annually. AI interviews from $520 a month. |
| Maki | Situational judgment questions about AI scenarios, and structured questions about how they’ve used AI before. | Not published. Credit-based, by quote. |
From each company’s own website, checked October 13, 2026. Prices and products change; tell us if something here is out of date and we’ll fix it.
Why the format matters
Asking someone how they’d check an AI’s work shows whether they know they should. Giving them a task where the AI’s first answer is wrong shows whether they do. SHOWCASE plants problems in the material, like an out-of-date spreadsheet or two notes that disagree, and the report shows whether the candidate found them, with links to the moments in the replay.
When another tool fits better
- You want cognitive, personality, or culture tests too. Bryq, TestGorilla, and Maki all offer them. SHOWCASE only tests how someone works with AI.
- You hire in high volume. A subscription with unlimited invitations can cost less per candidate. At $5 a test, SHOWCASE costs less than Bryq’s smallest annual plan up to about 165 tests a year.
- Your team works in ChatGPT or Gemini. SHOWCASE tests with Claude. The habits it measures carry over, but the tool won’t be the one they use at your company.
- You need an ATS integration. SHOWCASE doesn’t have one yet; you send candidates a link.
Questions
What’s the difference between AI skills tests?
Mainly whether the candidate works with an AI on a real task or answers questions about how they would. SHOWCASE has candidates do a task with Claude, with problems planted in the material, and records every message, command, and file. Bryq uses a short simulation, TestGorilla mostly multiple-choice tests and AI interviews, and Maki situational judgment questions.
Which AI skills assessment is cheapest?
For a few hires a year, SHOWCASE at $5 a test with no subscription. Bryq’s smallest annual plan is $828 a year with unlimited invitations, so it costs less per candidate above about 165 tests a year. TestGorilla starts at $215 a month billed annually, and Maki doesn’t publish prices. Checked October 13, 2026.
Is there an AI fluency test where candidates actually use AI?
Yes. In SHOWCASE, candidates use Claude in a chat or in Claude Code during a 20-minute task (45 for engineering), and the report links to the moments that show how they set it up, checked it, and corrected it.