# TestGorilla vs Codility: which do AI models recommend for technical hiring assessm, September 2026

HR AI Recommendation Index, September 2026 Edition, Technical hiring assessments. Zero of twelve models named TestGorilla first on the direct prompt; one named Codility. Page: https://hr-ai-index.com/talent/technical-hiring-assessments/testgorilla-vs-codility/

| | First-choice share | Rank | Negative rate | Labels | Models naming it |
|---|---|---|---|---|---|
| TestGorilla | 5% | #5 of 6 | 16% | 31 | 12 of 12 |
| Codility | 5% | #6 of 6 | 32% | 44 | 12 of 12 |

## The direct prompt, model by model

- Kimi K2: codility first (first choices: Codility) (alternatives: CodeSignal, Coderbyte, HackerRank, TestGorilla)
- GPT-5.4 mini: neither first, one named (first choices: HackerRank) (alternatives: CodeSignal, CoderPad, Codility)
- Gemini 3.5 Flash: neither first, one named (first choices: CoderPad) (alternatives: CodeSignal, HackerEarth, TestGorilla)
- DeepSeek V4 Flash: neither first, one named (first choices: CodeSignal) (alternatives: Codility, HackerRank, TestGorilla)
- MiniMax M2.5: neither first, one named (first choices: CodeSignal) (alternatives: Augment Code, HackerRank, TestGorilla)
- Claude Haiku 4.5: neither named
- Perplexity Sonar: neither named (first choices: HackerRank) (alternatives: CodeSignal, CoderPad, Coderbyte, Goodfit)
- Grok 4.1 Fast: neither named (first choices: HackerRank) (alternatives: CodeSignal, Coderbyte)
- Mistral Small: neither named (first choices: HackerRank) (alternatives: Augment Code, Coderbyte)
- Llama 4 Maverick: neither named (alternatives: CodeSignal, CoderPad, Coderbyte, Goodfit, HackerRank)
- Qwen 3.7 Flash: neither named (first choices: HackerRank) (alternatives: CodeSignal, TestDome, Vervoe)
- GLM 4.7 FlashX: neither named (first choices: HackerRank) (alternatives: CodeSignal)

## What the models said about TestGorilla

- "Billing issues: Multiple reviews mention unethical business practices, with charges continuing after cancellation" (MiniMax M2.5, negative prompt, hard negative)
- "It has a reputation for being buggy and confusing... making it a poor measure of senior engineering ability" (DeepSeek V4 Flash, negative prompt, hard negative)
- "Frequently called one of the "worst online assessment tools" due to poor usability, confusing interfaces" (Grok 4.1 Fast, negative prompt, hard negative)

## What the models said about Codility

- "it is often described as "quite pricey" ... Codility has had issues with "test integrity," where questions and answers are not sufficiently hidden from recruiters" (GLM 4.7 FlashX, negative prompt, hard negative)
- "Opaque scoring: Candidates and reviewers criticize edge-case scoring and lack of transparency" (MiniMax M2.5, negative prompt, hard negative)
- "Why we recommend avoiding HackerRank or Codility (for now)" (Gemini 3.5 Flash, paraphrase prompt, hard negative)
- "Codility offers the best balance of capability, reputation, and price" (Kimi K2, direct prompt, first choice)
- "Often ranked as the best overall, especially for enterprise use." (Mistral Small, comparative prompt, first choice)

Share is the count of first choices across the direct, paraphrase, budget and scale prompts over all twelve models, for a mid-market B2B company; rank is within the category. Comparisons are drawn for the top eight products in each category. Published under CC BY 4.0; the output is the models' output, and nothing here is a recommendation by the index.