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 ↑.
461/461
Filter models
No models match the current search and filters.
| Rank | Model | Company | Instructions following Score | Score | Total Cost | Tests Correct | Response Time (avg) |
|---|---|---|---|---|---|---|---|
| #113 | Kimi K3.1 Pro high | Moonshot AI | 1.0 | 0.8 | $1.646 | 2/2 | 5.71s |
| #114 | Seed-2.0-Code low | Bytedance Seed | 1.0 | 0.8 | $0.881 | 2/2 | 12.2s |
| #116 | Claude Opus 4.7 medium | Anthropic | 1.0 | 0.8 | $3.636 | 2/2 | 1.57s |
| #117 | DeepSeek V4 Flash 0423 high | DeepSeek | 1.0 | 0.8 | $0.261 | 2/2 | 15.4s |
| #118 | DeepSeek V4.1 Flash low | DeepSeek | 1.0 | 0.8 | $0.494 | 2/2 | 8.38s |
| #119 | GPT-5.6 Luna high | OpenAI | 1.0 | 0.8 | $0.118 | 2/2 | 1.79s |
| #120 | Nemotron 3 Ultra medium | NVIDIA | 1.0 | 0.8 | $0.674 | 2/2 | 6.35s |
| #121 | GPT-5.4 Nano medium | OpenAI | 1.0 | 0.8 | $0.235 | 2/2 | 1.88s |
| #122 | Gemini 3.8 Flash medium | 1.0 | 0.8 | $0.769 | 2/2 | 4.81s | |
| #123 | Qwen3.8 Flash Next low | Qwen | 1.0 | 0.8 | ~$0.045 | 2/2 | 2.32s |
| #124 | GPT-5.6 Terra low | OpenAI | 1.0 | 0.8 | $0.435 | 2/2 | 1.48s |
| #125 | Qwen3.5-27B medium | Qwen | 1.0 | 0.8 | $1.135 | 2/2 | 19.7s |
| #126 | GLM 5.3 FlashX max | Z.ai | 1.0 | 0.8 | $0.623 | 2/2 | 2.23s |
| #127 | Claude Fable 5.1 high | Anthropic | 1.0 | 0.8 | $3.471 | 2/2 | 5.60s |
| #128 | GPT-5.4 Mini medium | OpenAI | 1.0 | 0.8 | $1.102 | 2/2 | 2.13s |