Muse Wonder Full (Max)
MetaProprietary
描述
Muse Wonder Full is a proprietary reasoning model from Meta Superintelligence Labs, released on October 12, 2026 for personal and professional agents and long-running coding tasks. Active Memory retains relevant context across interactions, and Active Focus supports sustained task execution and error recovery. Muse Wonder Full is the higher-capability variant, available to Muse Code Power Usage and Maximum Usage subscribers. It is not available through the Meta Model API. The model card specifies text, image, video, and audio inputs, text output, and a 1M-token inference context for both variants.
发布日期
2026-10-12
参数规模
—
上下文长度
—
支持模态
—
能力雷达图
55
general
63
coding
50
reasoning
59
science
50
agents
50
multimodal
排行榜排名
基准测试分数 (LLM Stats)
(LLM Stats (zeroeval))Agents
MemoryBench
85.5%自报
JobBench
75.3%自报
Personal Superintelligence Index
59.1%自报
Program Bench
59.0%自报
FocusBench
54.2%自报
AutomationBench-AA
45.4%自报
Code
DeepSWE 1.1
79.0%自报
SWE-Marathon
57.5%自报
Terminal-Bench 4.0-AA
57.1%自报
Finance
Vals Finance Agent v2
59.1%自报
General
Artificial Analysis
56.0%自报
GDP.pdf
33.2%自报
Healthcare
HealthBench Professional
62.3%自报
Knowledge
Harvey LAB-AA
13.1%自报
Multimodal
OSWorld 2.1 (partial)
78.4%自报
OSWorld 2.1 (strict)
39.4%自报
AA 评测指数
(Artificial Analysis)Lcr(Artificial Analysis)83.7
Scicode(UIUC + Argonne National Lab (2024))62.6
Terminalbench V4 057.1
Intelligence Index(Artificial Analysis)55.8
Hle(Center for AI Safety + Scale AI (2025))53.6
LLM Stats 分类评分
(LLM Stats (zeroeval))Long Context90
Productivity80
Finance60
General60
Healthcare60
Agents60
Code60
Knowledge50
Multimodal50
Reasoning50
Tool Calling50
Vision50
定价
输入价格免费
输出价格免费
混合价格(3:1)免费
速度
Tokens/秒0.0
首Token延迟0.00s
首回答延迟0.00s
供应商价格排行
暂无提供商数据