Qwen3.8-Max
Alibaba Cloud · qwen3.8-max · qwen3.8-max-2026-09-02(快照 qwen3.8-max-0902) · 上线 ≈2026-09-03
用量热度
OpenRouter 用量榜的调用 token 量,按取数窗口标注日期。热度≠能力。
人类偏好
LMArena 人类偏好评测(Arena Score,text 口径),带票数与置信区间。
config: text_style_control · f566816dfe80
官方价格
厂商官方定价页的每百万 token 挂牌价,逐价带来源、检索日期、币种与计价条件。
官方文档按地域分区计价(USD);per_mtok 展示取 International/Singapore 档,不以更低档冒充通用价;Batch File 价仅北京区列示;官网注明不含限时促销。
| 分档 | in | out | cache | 条件 |
|---|---|---|---|---|
| Singapore / International | $2 | $6 | $0.25 | 显式缓存创建 $2.50;显式缓存读 $0.17 |
| China (Beijing) | $1.65 | $4.951 | $0.206 | 输入(Batch File) $0.825;输出(Batch File) $2.475;显式缓存创建 $2.063/读 $0.137 |
| Other regions (Frankfurt/US/Tokyo/HK, Global scope) | $1.65 | $4.951 | $0.206 |
来源: https://www.alibabacloud.com/help/en/model-studio/qwen3-8-max · 检索 2026-11-08
OpenRouter 挂牌价(市场参考,非厂商直连价): $2/$6
不含订阅套餐与企业协议价
上下文窗口
1,000,000 tokens · 来源
各来源名称对照(版本消歧)
不同版本与档位是独立行,绝不合并或平均。High/Max 档位来自 LMArena 展示名;厂商按模型 ID 定价,不对档位单独加价。
| source | 原始名 | 匹配 |
|---|---|---|
| openrouter-models | qwen/qwen3.8-max-0902 | auto/high |
| lmarena-hf-snapshot | Qwen3.8-Max | manual/high |
| vendor-official-pricing | qwen3.8-max | manual/high |
数据来源与署名
(本站本轮未接入该来源数据,尚无可标注的快照日期;下方引用为规定署名句式。) Source: OpenRouter (openrouter.ai/rankings), as of 未接入,无快照日期. Licensed under CC BY 4.0.
Data provided by OpenRouter under CC BY 4.0, AS IS; no warranties are made as to accuracy or completeness.
Adapted by AI Hot: token figures displayed as published; ordering and currency formatting applied by AI Hot.
Source: Arena Leaderboard Dataset (Arena Intelligence, Inc.), dataset lmarena-ai/leaderboard-dataset on Hugging Face, snapshot f566816dfe80f18d0bc804dd5deafa68d5957cb9, leaderboard published 2026-11-06. Licensed under CC BY 4.0. https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset
Data provided by Arena Intelligence, Inc. under CC BY 4.0, AS IS; no warranties are made as to accuracy or completeness.
Adapted by AI Hot: entries filtered to the categories shown; no scores altered.
AI Hot 汇总公开榜单数据并链回各官方来源。数值反映各来源自身口径,不构成合并或归一化排名。用量反映调用量而非模型能力。模型与名称归各自所有者。