Muse Spark

Meta·Released 2026-04-08

Open Source

Compare

Quality Index

52.1

5th of 469

Top 1%

Coding Index

47.5

10th of 379

Top 3%

Benchmarks

MMLU-ProNot evaluated

GPQA Diamond

88.4%

HLE

39.9%

LiveCodeBenchNot evaluated

SciCode

51.5%

TerminalBench Hard

45.5%

MATH-500Not evaluated

AIMENot evaluated

AIME 2025Not evaluated

IFBench

75.9%

Long Context Recall

69.7%

Tau2

91.5%

Market AverageTop Score

Open Source

Quick Compare

Similar Models

Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)

Anthropic

Q: 51.7$6.00/1M

Coding: +3.4

Z.ai: GLM 5.1

Z AI

Q: 51.4$1.94/1M203K ctx

Coding: -4.1

OpenAI: GPT-5.2

OpenAI

Q: 51.3$4.81/1M400K ctx

Claude Opus 4.6 (Adaptive Reasoning, Max Effort)

Anthropic

Q: 53.0$10.00/1M

OpenAI: GPT-5.3-Codex

OpenAI

Q: 53.6$4.81/1M400K ctx

Coding: +5.6

Qwen: Qwen3.6 Plus

Alibaba

Q: 50.0$0.73/1M1.0M ctx

Quality: -2.1Coding: -4.6

Compare all 7 models

Data Sources & Methodology

Artificial AnalysisSynced 2h ago

Benchmarks, quality index, speed metrics

For full methodology details, see our methodology page.