跳轉到主要內容

Muse Spark

MetaProprietary

描述

Muse Spark is the first model in the Muse family developed by Meta Superintelligence Labs. It is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration. It features a Contemplating mode that orchestrates multiple agents reasoning in parallel. It demonstrates competitive performance in multimodal perception, reasoning, health, and agentic tasks, with Contemplating mode achieving 58% on Humanity's Last Exam and 38% on FrontierScience Research.

發布日期
2026-04-08
參數規模
—
上下文長度
—
支援模態
—

能力雷達圖

70
general
70
coding
60
reasoning
77
science估算
80
agents
70
multimodal

缺少專門科學評測時,Science 由 LLM Stats 科學得分或推理能力估算。

排行榜排名

領域#排名分數來源
多模態榜22
63.0
LS

基準測試分數 (LLM Stats)

(LLM Stats (zeroeval))

Communication

Tau2 Telecom91.5%自報

Healthcare

MedXpertQA78.4%自報
HealthBench Hard42.8%自報

Reasoning

GPQANYU + Cohere + Anthropic (2023)89.5%自報
CharXiv-R86.4%自報
IPhO 202582.6%自報
LiveCodeBench Pro0.80 / 3000自報
SWE-Bench Verified77.4%自報
Terminal-Bench 2.0Stanford × Laude Institute (2026)59.0%自報
Humanity's Last Exam58.4%自報
SWE-Bench ProPrinceton NLP (2024)52.4%自報
ARC-AGI v242.5%自報
FrontierScience Research38.3%自報

Search

DeepSearchQA74.8%自報

Vision

ScreenSpot Pro84.1%自報
MMMU-Pro80.4%自報
SimpleVQA0.71 / 100自報
ERQA64.7%自報
ZEROBench0.33 / 100自報

AA 評測指數

(Artificial Analysis)

暫無 AA 評測資料

LLM Stats 分類評分

(LLM Stats (zeroeval))
Physics
90
Biology
90
Chemistry
90
Communication
90
Frontend Development
80
Grounding
80
Tool Calling
80
Image To Text
70
Multimodal
70
Reasoning
70
Search
70
General
70
Code
70
Vision
70
Math
60
Spatial Reasoning
60
Healthcare
60
Agents
60
Science
40

定價

暫無定價資料

速度

暫無速度資料

供應商價格排行

暫無提供商資料

外部連結