Benchmark
WildBench: haiku_reward.Planning & Reasoning.K=-1
People compare model answers and choose which they prefer. Higher scores are better.
- Models measured
- 32
- Scores recorded
- 32
- Sources
- 1
- IQ weight
- 1.40
Score distribution
- 3.8
- Mean
- 10.4
- Median
- 32.6
- Std dev
- -32.2
- 10th percentile
- 45.2
- 90th percentile
- -75.7
- Min
This benchmark has a weight of 1.40 for chat models, measured over 32 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | GPT 4oallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 53.1 |
| 2 | GPT-4 Turboallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 46.6 |
| 3 | Claude 3.5 Sonnetallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 46.4 |
| 4 | Gemini 1.5 Proallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 45.2 |
| 5 | GPT-4-0125-previewallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 40.3 |
| 6 | Claude 3 Opusallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 32.9 |
| 7 | Llama 3 70B Instructallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 32.8 |
| 8 | Gemini 1.5 Flashallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 32.3 |
| 9 | yi-largeallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 29.4 |
| 10 | Reka Coreallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 28.3 |
| 11 | Yi 1.5 34B Chatallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 27.7 |
| 12 | Nemotron-4 340B Instructallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 27.4 |
| 13 | Claude 3 Sonnetallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 22.5 |
| 14 | DeepSeek Coder V2 Instructallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 20.8 |
| 15 | Qwen2 72B Instructallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 13.2 |
| 16 | GLM-4-9b-chatallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 10.4 |
| 17 | Command R Plusallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | 8.4 |
| 18 | Starling LM 7B Betaallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | -7.3 |
| 19 | Mistral Largeallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307, 2402· WildBench direct data | -7.4 |
| 20 | Reka Flashallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | -7.8 |
| 21 | Llama 3 8B Instructallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | -9.7 |
| 22 | Command R+allenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | -13 |
| 23 | Mixtral 8x7B Instruct v0.1allenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | -17.9 |
| 24 | Nous Hermes 2 Mixtral 8x7B DPOallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | -20.5 |
| 25 | Tulu 2 DPO 70Ballenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | -23.5 |
| 26 | Reka Edgeallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | -24.3 |
| 27 | Llama-2-70b-chat-hfallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | -25.4 |
| 28 | Mistral 7B Instruct v0.2allenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | -27.3 |
| 29 | GPT-3.5 Turboallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | -32.2 |
| 30 | Llama-2-7b-chat-hfallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | -47.3 |
| 31 | Gemma 7B ITallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | -57.5 |
| 32 | Gemma 2B ITallenai/WildBench, haiku_reward.Planning & Reasoning.K=-1, claude-3-haiku-20240307· WildBench direct data | -75.7 |
Data credits: WildBench, Allen Institute for AI