Muse Spark
MetaProprietary
Description
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.
Release Date
2026-04-08
Parameters
—
Context Length
—
Modalities
—
Capability Radar
70
general
70
coding
60
reasoning
77
scienceest.
80
agents
70
multimodal
Science is estimated from LLM Stats science scores or reasoning when no dedicated science benchmarks are available.
Rankings
| Domain | #Rank | Score | Source |
|---|---|---|---|
| Multimodal Ranking | 22 | 63.0 | LS |
Benchmark Scores (LLM Stats)
(LLM Stats (zeroeval))Communication
Tau2 Telecom
91.5%SR
Healthcare
MedXpertQA
78.4%SR
HealthBench Hard
42.8%SR
Reasoning
GPQANYU + Cohere + Anthropic (2023)
89.5%SR
CharXiv-R
86.4%SR
IPhO 2025
82.6%SR
LiveCodeBench Pro
0.80 / 3000SR
SWE-Bench Verified
77.4%SR
Terminal-Bench 2.0Stanford × Laude Institute (2026)
59.0%SR
Humanity's Last Exam
58.4%SR
SWE-Bench ProPrinceton NLP (2024)
52.4%SR
ARC-AGI v2
42.5%SR
FrontierScience Research
38.3%SR
Search
DeepSearchQA
74.8%SR
Vision
ScreenSpot Pro
84.1%SR
MMMU-Pro
80.4%SR
SimpleVQA
0.71 / 100SR
ERQA
64.7%SR
ZEROBench
0.33 / 100SR
AA Evaluation Indices
(Artificial Analysis)No AA evaluation data available
LLM Stats Category Scores
(LLM Stats (zeroeval))Physics90
Biology90
Chemistry90
Communication90
Frontend Development80
Grounding80
Tool Calling80
Image To Text70
Multimodal70
Reasoning70
Search70
General70
Code70
Vision70
Math60
Spatial Reasoning60
Healthcare60
Agents60
Science40
Pricing
No pricing data available
Speed
No speed data available
Provider Price Ranking
No provider data available