AI BENCHY Category
Trivia Ranking
See which AI models perform best on Trivia, which ones stay reliable, and where the biggest gaps appear. Sort by: Tests Correct ↓.
Failure Reasons
| Rank | Model | Company | Trivia Score | Score | Tests Correct | Response Time (avg) |
|---|---|---|---|---|---|---|
| #98 | GPT-5.4 none | OpenAI | 3.0 | 5.7 | 0/1 | 990ms |
| #99 | gpt-oss-120b medium | OpenAI | 3.0 | 5.7 | 0/1 | 26.5s |
| #100 | Kimi K2.6 none | Moonshot AI | 3.0 | 5.7 | 0/1 | 1.36s |
| #101 | GLM 5.1 none | Z.ai | 3.0 | 5.7 | 0/1 | 2.34s |
| #102 | MiMo-V2.5-Pro none | Xiaomi | 3.0 | 5.7 | 0/1 | 1.89s |
| #104 | DeepSeek V3.2 none | DeepSeek | 3.0 | 5.7 | 0/1 | 17.2s |
| #105 | Qwen3.6 Flash none | Qwen | 3.0 | 5.6 | 0/1 | 649ms |
| #106 | MiniMax M2.5 medium | Minimax | 3.0 | 5.5 | 0/1 | 80.8s |
| #107 | Mistral Small 4 medium | Mistral | 3.0 | 5.5 | 0/1 | 5.92s |
| #109 | Elephant Alpha medium | Openrouter | 0.0 | 5.5 | 0/0 | 0ms |
| #110 | Qwen3.5-122B-A10B none | Qwen | 3.0 | 5.5 | 0/1 | 295ms |
| #111 | Nemotron 3 Nano Omni 30b A3b Reasoning medium | NVIDIA | 0.0 | 5.4 | 0/0 | 0ms |
| #112 | Kimi K2.5 none | Moonshot AI | 3.0 | 5.4 | 0/1 | 3.90s |
| #114 | GLM 5 Turbo none | Z.ai | 3.0 | 5.4 | 0/1 | 2.37s |
| #115 | Laguna M.1 none | Poolside | 0.0 | 5.4 | 0/0 | 0ms |