Qwen3-Next

Benchmark Model Rank Results
code-generation-on-aider-polyglotQwen3-Next-80B-A3B-Instruct–Accuracy: 49.8
code-generation-on-livecodebench-v6Qwen3-Next-80B-A3B-Instruct–Pass@1: 56.6
code-generation-on-multipl-eQwen3-Next-80B-A3B-Instruct–Accuracy: 87.8
instruction-following-on-arena-hard-auto-v2Qwen3-Next-80B-A3B-Instruct–Accuracy: 82.7
multi-task-language-understanding-on-includeQwen3-Next-80B-A3B-Instruct–Accuracy: 78.9
multi-task-language-understanding-on-livebench-2024-11-25Qwen3-Next-80B-A3B-Instruct–Score: 75.8
multi-task-language-understanding-on-mmlu-proQwen3-Next-80B-A3B-Instruct–Accuracy: 80.6
multi-task-language-understanding-on-mmlu-reduxQwen3-Next-80B-A3B-Instruct–Accuracy: 90.9
multiple-choice-question-answering-mcqa-on-gpqa-diamondQwen3-Next-80B-A3B-Instruct–Accuracy: 72.9
multiple-choice-question-answering-mcqa-on-supergpqaQwen3-Next-80B-A3B-Instruct–Accuracy: 58.8
text-generation-on-writingbenchQwen3-Next-80B-A3B-Instruct–Score (0-10): 8.73