AI BENCHY Category
Instructions following Ranking
See which AI models perform best on Instructions following, which ones stay reliable, and where the biggest gaps appear. Sort by: Tests Correct ↑.
| Rank | Model | Company | Instructions following Score | Score | Tests Correct | Response Time (avg) |
|---|---|---|---|---|---|---|
| #12 | Gemini 3.1 Flash Lite Preview high | 9.8 | 8.6 | 2/2 | 64.0s | |
| #13 | Grok 4.20 Beta medium | X AI | 9.8 | 8.5 | 2/2 | 4.89s |
| #14 | Qwen3.6 Max Preview medium | Qwen | 10.0 | 8.5 | 2/2 | 24.3s |
| #15 | GPT-5.3-Codex medium | OpenAI | 10.0 | 8.4 | 2/2 | 3.04s |
| #16 | Gemini 3 Flash Preview low | 9.9 | 8.4 | 2/2 | 7.02s | |
| #17 | GLM 5 medium | Z.ai | 10.0 | 8.3 | 2/2 | 7.25s |
| #18 | Qwen3.7 Plus medium | Qwen | 10.0 | 8.2 | 2/2 | 16.1s |
| #19 | Seed-2.0-Lite medium | Bytedance Seed | 10.0 | 8.2 | 2/2 | 7.26s |
| #20 | Gemini 3.5 Flash none | 9.8 | 8.1 | 2/2 | 3.38s | |
| #21 | GPT-5.4 medium | OpenAI | 10.0 | 8.0 | 2/2 | 3.11s |
| #22 | Step 3.7 Flash medium | Stepfun | 9.8 | 8.0 | 2/2 | 1.83s |
| #23 | GLM 5 Turbo medium | Z.ai | 10.0 | 8.0 | 2/2 | 5.38s |
| #24 | GPT-5.2 Chat none | OpenAI | 9.8 | 7.9 | 2/2 | 5.51s |
| #25 | Qwen3.5 Plus 2026-02-15 medium | Qwen | 10.0 | 7.9 | 2/2 | 31.9s |
| #26 | Qwen3.6 Plus medium | Qwen | 10.0 | 7.9 | 2/2 | 7.54s |