Writing on Arena-Write v1 (test)
1,447EloLongWriter-Zero
Evaluation results
| Method | Links | |
|---|---|---|
| LongWriter-ZeroModel Scale=32B, Continual Pretrain=true, Thinking=true2025.06 | 1,447 | |
| DeepSeek-R1Provider=DeepSeek-AI2025.06 | 1,343 | |
| Qwen3-235B-A22BModel Scale=235B-A22B2025.06 | 1,343 | |
| DeepSeek-V3Provider=DeepSeek-AI2025.06 | 1,236 | |
| LongWriter-ZeroModel Scale=32B, Continual Pretrain=false, Thinking=true2025.06 | 1,221 | |
| Claude-Sonnet-4Provider=Anthropic2025.06 | 1,185 | |
| o1-PreviewProvider=OpenAI2025.06 | 1,080 | |
| Qwen2.5-MaxProvider=Alibaba2025.06 | 1,029 | |
| GPT-4o-2024-11-20Provider=OpenAI2025.06 | 947 | |
| Qwen-2.5-72B-InstructModel Scale=72B2025.06 | 911 | |
| Mistral-Large-InstructProvider=Mistral2025.06 | 724 | |
| LongWriter-ZeroModel Scale=32B, Continual Pretrain=true, Thinking=false2025.06 | 668 | |
| Qwen-2.5-7B-InstructModel Scale=7B2025.06 | 661 | |
| Llama-3.3-70B-InstructModel Scale=70B2025.06 | 570 | |
| LongWriterModel Scale=8B2025.06 | 457 | |
| Llama-3.1-8B-InstructModel Scale=8B2025.06 | 445 |