AI BENCHY 失敗分析
不正解 の失敗
どのAIモデルで 不正解 が起きやすいかを確認し、選ぶ前に信頼性のリスクを見極められます。 並び替え: 失敗数 ↑.
| 順位 | モデル | 企業 | 不正解 件数 | スコア | 正解テスト | 応答時間（平均） |
|---|---|---|---|---|---|---|
| #43 | Qwen3.5-35B-A3B medium | Qwen | 2 | 7.4 | 10/18 | 44.5s |
| #5 | Gemini 3 Flash Preview low | 3 | 8.8 | 15/18 | 6.01s | |
| #6 | Seed-2.0-Lite medium | Bytedance Seed | 3 | 8.6 | 13/18 | 30.4s |
| #7 | GPT-5.3-Codex medium | OpenAI | 3 | 8.6 | 13/18 | 15.4s |
| #9 | Qwen3.6 Plus Preview medium | Qwen | 3 | 8.5 | 13/17 | 13.9s |
| #11 | Gemini 3.1 Flash Lite Preview high | 3 | 8.4 | 12/16 | 68.8s | |
| #12 | Gemini 3 PRO Preview medium | 3 | 8.4 | 14/18 | 9.06s | |
| #16 | GPT-5.4 medium | OpenAI | 3 | 8.2 | 13/18 | 18.6s |
| #18 | GLM 5 Turbo medium | Z.ai | 3 | 8.1 | 12/18 | 17.7s |
| #19 | Qwen3.5-122B-A10B medium | Qwen | 3 | 8.1 | 13/18 | 31.4s |
| #20 | Qwen3.6 Plus medium | Qwen | 3 | 8.1 | 13/18 | 15.3s |
| #23 | MiMo-V2-Pro medium | Xiaomi | 3 | 8.1 | 12/18 | 12.3s |
| #25 | Grok 4.20 Beta medium | X AI | 3 | 8.0 | 12/18 | 9.81s |
| #27 | DeepSeek V3.2 medium | DeepSeek | 3 | 8.0 | 12/18 | 46.4s |
| #30 | Step 3.5 Flash medium | Stepfun | 3 | 7.9 | 11/17 | 26.8s |