Meta

Muse Spark Open

by Metatext+image+file+audio+video->text5 endpoints
Neura Intelligence Index
86.5/ 100
Rank
#24—
Confidence
low
Benchmarks
3(8% coverage)

Overview

Muse Spark Open is a coding-optimized open-weight model. Trained specifically for agentic coding workflows, it delivers higher first-attempt accuracy, reliable tool calling, and deep codebase understanding across the repos developers actually work in.

Capabilities

Text generation
Image understanding
Audio processing
Video understanding
Tool use / Function calling
Structured output
Extended reasoning
Prompt caching
Long context
File input
Top-P sampling
Stop sequences
Deterministic seed
Log probabilities

Modalities

Input
TextImageVideoFileAudio
Output
Text

Technical Specifications

Context Window
1.0M tokens
Max Output
943.7K tokens
Tokenizer
Other
Uptime (24h)
10000.0%
Parameters
816000000000B
License
muse_model_licenseOpen Weights
Country
🇺🇸US
Release Date
October 15, 20261w

Supported Parameters

frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_ppresence_penaltyreasoningreasoning_effortrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p

Benchmark Performance

coding

BenchmarkScoreSource
SciCode
58.5
verified

reasoning

BenchmarkScoreSource
GPQA Diamond
89
verified

agent

BenchmarkScoreSource
MCP Atlas
89.2
verified

Performance Over Time

Pricing

Per 1M tokens
Input$1.25
Output$4.25
Cache read$0.15
Added
October 15, 2026
Last synced 10/23/2026