---
title: "Muse Spark Benchmarks - RankedAGI"
description: "Detailed benchmark and metadata record for Muse Spark."
source: "https://rankedagi.com/models/muse-spark"
---

# Muse Spark

| Field | Value |
| --- | --- |
| Organization | Meta |
| License | Proprietary |
| Type | Thinking |
| Version | Latest |
| Released | 2026-04-08 |
| Last updated | 2026-05-04T03:43:08.354Z |

## Benchmarks

| Benchmark | Category | Value | Description | Source |
| --- | --- | --- | --- | --- |
| RankedAGI Coding | coding | 77.4% | RankedAGI Coding Score |  |
| SWEBench Pro | coding | 52.4% | Diverse Agentic Coding |  |
| SWEBench Verified | coding | 77.4% | Agentic Coding | https://www.swebench.com/index.html |
| Terminal-Bench 2.0 | coding | 59.0% | Agentic Terminal Coding | https://www.tbench.ai/leaderboard/terminal-bench/2.0 |
| RankedAGI Agentic | Agents | 72.0% | RankedAGI Agentic Score |  |
| Code Arena | coding | 1510 | Code Arena Elo Score | https://arena.ai/leaderboard/code |
| CyberGym | safety | 43.5% |  |  |
| RankedAGI Reasoning | reasoning | 69.9% | RankedAGI Reasoning Score |  |
| Humanity's Last Exam | reasoning | 42.8% | Multidisciplinary Reasoning (no tools) |  |
| Humanity's Last Exam w/ Tools | reasoning | 50.4% | Multidisciplinary Reasoning (with tools) |  |
| GPQA Diamond | reasoning | 89.5% | Generalized Prefix Question Answering Score (Reasoning) PhD Level Reasoning |  |
| MMMU-Pro | imaging | 80.4% | Multimodal Understanding |  |
| RankedAGI Math | math | 72.5% | RankedAGI Math Score |  |
| RankedAGI Overall | general | 66.3% | Overall RankedAGI score |  |
| ARC AGI 2 | reasoning | 42.5% | Abstract Reasoning Puzzles (Public) |  |
| LiveCodeBench Pro | coding | 80.0% | Competitive Coding |  |
| 𝜏²-Bench Telecom | Agents | 91.5% | Agentic Tool Use |  |
| HealthBench Hard | Health | 42.8% | Open-ended Health Queries |  |
| MedXpertQA Text | Health | 52.6% | Medical Multiple Choice |  |
| MedXpertQA (MM) | Health | 78.4% | Medical Multiple Choice |  |
| DeepSearch QA | Agents | 74.8% | Agentic Search |  |
| GDPval-AA Elo | Agents | 1428 | Office Tasks (Artificial Analysis) |  |
| ZeroBench | general | 33.0% | Multi-step Visual Reasoning |  |
| CharXiv Reasoning | Multimodal | 86.4% | Figure Understanding |  |
| Code DesignArena | Design | 1318 |  | https://www.designarena.ai/leaderboard/code |

## Model Links

| Title | URL |
| --- | --- |
| Release | https://ai.meta.com/blog/introducing-muse-spark-msl/ |