Back to BenchmarksMedium
TriviaQA
Trivia questions requiring broad knowledge
Leaderboard
(18 models)| Rank | Model | Score | Stderr |
|---|---|---|---|
| 1 | Llama-2-70b-hf | 87.60 | — |
| 2 | Claude 3.7 Sonnet | 87.50 | — |
| 3 | Llama 3.1 405B | 86.00 | — |
| 4 | gpt-3.5-turbo-1106 | 85.80 | — |
| 5 | GPT-4.1 | 84.80 | — |
| 6 | Llama-2-7b | 84.60 | — |
| 7 | DeepSeek V3 | 82.90 | — |
| 8 | Mixtral-8x7B-v0.1 | 82.20 | — |
| 9 | falcon-180B | 79.90 | — |
| 10 | Claude Opus 4.5 | 78.90 | — |
| 11 | Mistral-7B-v0.1 | 75.20 | — |
| 12 | Phi-3-medium-128k-instruct | 73.90 | — |
| 13 | gemma-7b | 72.30 | — |
| 14 | Qwen 2.5 72B | 71.90 | — |
| 15 | Meta-Llama-3-8B-Instruct | 67.70 | — |
| 16 | Phi-3-mini-4k-instruct | 64.00 | — |
| 17 | Phi-3-small-8k-instruct | 58.10 | — |
| 18 | Phi-4 | 45.20 | — |
Data source: Epoch AI, “Data on AI Benchmarking”. Published at epoch.ai
Licensed under CC-BY 4.0