← sources

Artificial Analysis updated Qwen 3.8 Flash Next score as 29

source: reddit (reddit:LocalLLaMA)published 2026-11-13 04:06ingested 2026-11-14 02:42

retained text

Between GPT-6 Luna Max (28) and Claude Haiku 5.5 High (30).  Comparing to other Qwen 3.8, the model is closer to max (34) than to 27B (20).

[https://artificialanalysis.ai/models/qwen3-8-flash-next](https://artificialanalysis.ai/models/qwen3-8-flash-next)

document fetched 2026-11-14 00:46 · canonical https://reddit.com/r/LocalLLaMA/comments/1xuxzfs/artificial_analysis_updated_qwen_38_flash_next