AI BENCHY Category Failures
Domain specific: Wrong answer
Domain specific
Wrong answer
See which AI models are most likely to hit Wrong answer on Domain specific, so you can spot weak points faster.
Failure Reasons
| Rank | Model | Company | Wrong answer Count | Category Score | Tests Correct | Response Time (avg) |
|---|---|---|---|---|---|---|
| #38 | Grok 4.3 medium | X AI | 2 | 5.3 | 1/3 | 181.7s |
| #46 | Qwen3.6 35B A3B medium | Qwen | 2 | 5.3 | 1/3 | 22.5s |
| #50 | Gemini 3.1 Flash Lite Preview low | 2 | 5.3 | 1/3 | 2.36s | |
| #54 | GPT-5 Mini medium | OpenAI | 2 | 3.6 | 0/3 | 44.6s |
| #57 | Step 3.7 Flash low | Stepfun | 2 | 5.3 | 1/3 | 43.3s |
| #58 | Gemini 3.1 Flash Lite Preview none | 2 | 5.3 | 1/3 | 942ms | |
| #59 | GLM 5V Turbo medium | Z.ai | 2 | 5.3 | 1/3 | 38.1s |
| #61 | Gemini 3.1 Flash Lite low | 2 | 5.3 | 1/3 | 1.52s | |
| #62 | Step 3.5 Flash medium | Stepfun | 2 | 5.3 | 1/3 | 170.5s |
| #64 | MiMo-V2-Flash medium | Xiaomi | 2 | 5.9 | 1/3 | 96.0s |
| #68 | Claude Opus 4.8 none | Anthropic | 2 | 5.3 | 1/3 | 1.66s |
| #70 | GPT-5.4 Nano medium | OpenAI | 2 | 5.9 | 1/3 | 38.2s |
| #71 | Step 3.7 Flash high | Stepfun | 2 | 4.1 | 0/3 | 149.6s |
| #72 | DeepSeek V3.2 medium | DeepSeek | 2 | 2.9 | 0/3 | 24.3s |
| #76 | Kimi K2.5 medium | Moonshot AI | 2 | 3.5 | 0/3 | 137.3s |