Home Model series Qwen Next Qwen3-Next-80B-A3B-Thinking Qwen3-Next-80B-A3B-Thinking · Arena-Hard v2 · 2 · GPT-4.1 judge · Unit not specified · Scores and configurations Qwen3-Next-80B-A3B-Thinking · Arena-Hard v2 · 2 · GPT-4.1 judge · Unit not specified · Scores and configurations Qwen3-Next-80B-A3B-Thinking record for Arena-Hard v2 · 2 · GPT-4.1 judge · Unit not specified. 1 score record for this release, including 1 exact value (may include historical records). Includes 1 source location.
Scores and evaluation conditions Qwen3-Next-80B-A3B-Thinking · Arena-Hard v2 scores and configurations Metric and score Configuration and evaluation conditions Sources Reported score 62.3 Reported 62.3 · Unit not specified Published Unconfirmed
Qwen3-Next-80B-A3B-Thinking Settings Official Qwen3-Next 80B total / 3B active open-weight variant; the cited model card limits this checkpoint to its named mode. Unknowns The performance table does not state per-row decoding, tools, or serving implementation.、Performance表与比较图未标明单位;旧结果百分比标签已撤回。 Tools Tool access not recorded Reasoning budget not recorded Framework not recorded Protocol GPT-4.1 judge Conditions 当前模型卡明确本权重仅支持thinking;表中未注明单位、实际逐项解码设置或完整harness。最佳实践设置不自动视为实测配置。 View this score in rankings
Benchmark versions and metrics Aliases Arena-Hard v2
Reported score · Unit not specified · Ranking direction unconfirmed