# Semgrep vs CodeQL: which do AI models recommend for SAST, October 2026

IT AI Recommendation Index, October 2026 Edition, Static application security testing. Four of fourteen models named Semgrep first on the direct prompt; zero named CodeQL. Page: https://it-ai-index.com/developer/static-application-security-testing/semgrep-vs-codeql/

| | First-choice share | Rank | Negative rate | Labels | Models naming it |
|---|---|---|---|---|---|
| Semgrep | 36% | #1 of 10 | 5% | 63 | 14 of 14 |
| CodeQL | 4% | #5 of 10 | 11% | 27 | 11 of 14 |

## The direct prompt, model by model

- GPT-5.4 mini: semgrep first (first choices: Semgrep) (alternatives: Checkmarx One, Snyk Code, Veracode Static Analysis)
- Perplexity Sonar: semgrep first (first choices: Semgrep) (alternatives: Checkmarx One, Snyk Code)
- Kimi K2: semgrep first (first choices: Semgrep, Snyk Code) (alternatives: SonarQube)
- GPT-6 Luna: semgrep first (first choices: Semgrep) (alternatives: GitHub Advanced Security, Snyk)
- Gemini 3.5 Flash: neither first, one named (first choices: Aikido Security) (alternatives: GitHub Advanced Security, GitLab SAST, Semgrep, Snyk Code)
- Grok 4.1 Fast: neither first, one named (first choices: Snyk Code) (alternatives: Semgrep, SonarQube)
- Mistral Small: neither first, one named (first choices: Snyk) (alternatives: Aikido Security, Semgrep)
- DeepSeek V4 Flash: neither first, one named (first choices: Snyk Code) (alternatives: CodeAnt.ai, Semgrep, SonarQube)
- Qwen 3.7 Flash: neither first, one named (first choices: SonarQube) (alternatives: Semgrep, Snyk Code)
- GLM 4.7 FlashX: neither first, one named (first choices: SonarQube) (alternatives: Checkmarx One, Codacy, DeepSource, GitHub Advanced Security, OpenText Fortify, Semgrep, Snyk Code, Veracode Static Analysis)
- MiniMax M2.5: neither first, one named (first choices: Snyk) (alternatives: Semgrep, Veracode Static Analysis)
- Muse Glimmer 30B: neither first, one named (first choices: SonarQube) (alternatives: Semgrep)
- Claude Haiku 4.5: neither named
- Llama 4 Maverick: neither named (first choices: Snyk Code)

## What the models said about Semgrep

- "Semgrep OSS and Bandit are fast and low false-positive for obvious patterns, but miss vulnerabilities that span multiple functions... insufficient as sole nightly security scanners" (Muse Glimmer 30B, negative prompt, soft negative)
- "Semgrep Community Edition as your only deep security scanner: It can be useful for fast checks and custom rules, but its analysis is limited to a file/function context" (GPT-6 Luna, negative prompt, soft negative)
- "Checkmarx, Veracode, and Semgrep have all struggled with this because their rule-based engines flag anything that could be a vulnerability" (Claude Haiku 4.5, negative prompt, soft negative)
- "Semgrep as the primary scanner across all languages... For most budget-constrained companies, Semgrep OSS + SonarQube Community Edition... is the most commonly recommended starting point" (Muse Glimmer 30B, budget prompt, first choice)
- "Tools like Semgrep are generally preferred for modern CI/CD because they are lightweight, customizable, and run significantly faster." (GLM 4.7 FlashX, negative prompt, first choice)
- "look for developer-friendly alternatives like Semgrep, where rules are written in YAML and read like standard code" (Gemini 3.5 Flash, negative prompt, first choice)

## What the models said about CodeQL

- "CodeQL requires a subscription to GitHub Advanced Security (GHAS), which is highly enterprise-focused and expensive" (Gemini 3.5 Flash, budget prompt, soft negative)
- "CodeQL (A "Proceed with Caution" for Closed-Source Teams) ... GHAS licensing, which can be prohibitively expensive" (Gemini 3.5 Flash, negative prompt, soft negative)
- "CodeQL if your stack or plan doesn't fit: Check language support before adopting it" (GPT-6 Luna, negative prompt, soft negative)
- "GitHub CodeQL – Free for public repositories ... It provides deep code analysis and is widely used for finding vulnerabilities in multiple languages." (Mistral Small, budget prompt, first choice)
- "my default recommendation would be GitHub CodeQL if you already use GitHub" (GPT-5.4 mini, paraphrase prompt, first choice)
- "Lower noise: CodeQL (semantic, low FPs)" (Grok 4.1 Fast, negative prompt, first choice)

Share is the count of first choices across the direct, paraphrase, budget and scale prompts over all fourteen 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.