跳轉到主要內容

Muse Spark Open (Xhigh)

Meta開源權重Muse Model License · 商用許可

描述

Muse Spark Open is Meta Superintelligence Labs' open-weight natively multimodal Mixture-of-Experts model (816B total / 42B active parameters) with a 1,048,576-token context window. The HF model card specifies text, image, video, and audio inputs and text output. It supports configurable reasoning_effort (low/medium/high/xhigh; default xhigh) and is positioned for coding and agentic workflows. Weights are released under the Muse Model License with limited commercial terms and are available on Hugging Face as meta-models/Muse-Spark-Open-816B-A42B (BF16) and an NVFP4 sibling. Hosted inference is offered by third-party providers including Together AI, Fireworks AI, Groq, Modal, Nebius, Baseten, and OpenRouter (slug meta/muse-spark-open).

發布日期
2026-10-15
參數規模
816.0B
上下文長度
1.0M
支援模態
audio, image, pdf, text, video

能力雷達圖

40
general
58
coding
89
reasoning
69
science
80
agents
70
multimodal

排行榜排名

領域#排名分數來源
智慧體能力模型榜81
44.0
LS
程式碼能力榜44
88.0
AA
通用能力榜77
66.0
AA
多模態榜23
62.0
LS
推理能力44
74.0
LB
科學能力26
83.0
AA

基準測試分數 (LLM Stats)

(LLM Stats (zeroeval))

Agents

Toolathlon-Verified72.8%自報
Job Bench61.8%自報

Code

DeepSWE 1.160.0%自報
SWE Atlas - Codebase QnA58.0%自報

General

GDPval-AA1401.00 / 3000自報
Artificial Analysis39.0%自報

Knowledge

AA-Omniscience Index25.00 / 200自報

Multimodal

OSWorld-Verified80.4%自報

Reasoning

MCP Atlas89.2%自報
GPQANYU + Cohere + Anthropic (2023)89.0%自報
Terminal-Bench 2.176.4%自報
Humanity's Last Exam (with tools)61.0%自報
SciCode58.5%自報

Vision

BabyVision80.2%自報
SimpleVQA0.74 / 100自報
ZEROBench0.50 / 100自報
WorldVQA49.4%自報

AA 評測指數

(Artificial Analysis)
Gpqa(NYU + Cohere + Anthropic (2023))
89.0
Lcr(Artificial Analysis)
78.3
Terminalbench V2 1
76.4
Scicode(UIUC + Argonne National Lab (2024))
58.4
Hle(Center for AI Safety + Scale AI (2025))
43.9
Intelligence Index(Artificial Analysis)
38.9
Terminalbench V4 0
8.1

LLM Stats 分類評分

(LLM Stats (zeroeval))
Legal
100
Finance
100
Agents
100
Reasoning
100
General
85
Knowledge
25
Science
25
Tool Calling
80
Image To Text
70
Multimodal
70
Physics
70
Biology
70
Chemistry
70
Code
70
Vision
70
Math
60
Productivity
60

定價

輸入價格$1.25 / 1M tokens
輸出價格$4.25 / 1M tokens
混合價格(3:1)$2 / 1M tokens
快取讀取價格$0.15 / 1M tokens

速度

Tokens/秒161.8
首Token延遲0.56s
首回答延遲12.93s

供應商價格排行

供應商價格排行

12 個供應商

最便宜: Together最貴: Venice AI
供應商輸入輸出
1Together最便宜
$0
$0
2Meta主要
$1.25
$4.25
3NanoGPT
$1.25
$4.25
4OpenRouter
$1.25
$4.25
5Kilo Gateway
$1.25
$4.25
6Cloudflare AI Gateway
$1.25
$4.25
7Ollama Cloud
$1.25
$4.25
8Requesty
$1.25
$4.25
9Vercel AI Gateway
$1.25
$4.25
10Merge Gateway
$1.25
$4.25
11Prime Inference
$1.25
$4.25
12Venice AI
$1.5625
$5.3125

比較該模型在不同 API 供應商之間的定價。

外部連結