DongfuJiang's picture
Publish task/model/context tree; migrate current Harbor results and all notes; current scoring protocol only
0054129 verified
|
Raw History Blame Contribute Delete
1.51 kB
# LeetCode
| Model | Trace | Website | Current published fractions by context |
|---|---|---|---|
| [Claude Opus 5.5](https://huggingface.co/datasets/TIGER-Lab/WebsiteBenchArtifacts/tree/main/benchmark/leetcode/claude-opus-5-5) | ✓ | ✓ | 1m: 242/326 |
| [Claude Opus 5](https://huggingface.co/datasets/TIGER-Lab/WebsiteBenchArtifacts/tree/main/benchmark/leetcode/claude-opus-5) | ✓ | ✓ | 1m: 62/326 |
| [GPT 5.6 Luna](https://huggingface.co/datasets/TIGER-Lab/WebsiteBenchArtifacts/tree/main/benchmark/leetcode/gpt-5.6-luna) | ✓ | ✓ | recorded: 10/326 |
| [Grok 4.6](https://huggingface.co/datasets/TIGER-Lab/WebsiteBenchArtifacts/tree/main/benchmark/leetcode/grok-4.6) | ✓ | ✓ | recorded: 26/326 |
| [GLM 5.3 Flash](https://huggingface.co/datasets/TIGER-Lab/WebsiteBenchArtifacts/tree/main/benchmark/leetcode/glm-5.3-flash) | ✓ | ✓ | 1m: 30/326; 200k: 45/326 |
| [Kimi K3](https://huggingface.co/datasets/TIGER-Lab/WebsiteBenchArtifacts/tree/main/benchmark/leetcode/kimi-k3) | ✓ | ✓ | 200k: 77/326 |
| [DeepSeek V4.1 Flash](https://huggingface.co/datasets/TIGER-Lab/WebsiteBenchArtifacts/tree/main/benchmark/leetcode/deepseek-v4.1-flash) | ✓ | ✓ | 1m: 73/326 |
| [Qwen 3.8 Max](https://huggingface.co/datasets/TIGER-Lab/WebsiteBenchArtifacts/tree/main/benchmark/leetcode/qwen3.8-max) | ✓ | ✓ | 1m: 60/326; 200k: 80/326 |
| [GPT 6.1 Sol](https://huggingface.co/datasets/TIGER-Lab/WebsiteBenchArtifacts/tree/main/benchmark/leetcode/gpt-6.1-sol) | ✓ | ✓ | 258k: 228/326 |