Request a pilot
‹ PublicAI Index

The LLM benchmark aggregator.

Muse Spark Open

Meta · 822B · open weights

Strongest in Web research (#15 of 49, on 1 of its 2 boards), weakest in Core abilities (#149 of 225). Above par in 4 of 14 scopes. Among the models it meets almost everywhere, it finishes behind Claude Fable 5 and GPT-6 Astra and ahead of GPT-5 mini and DeepSeek V3.2.

§ 1 · Profile

What it is good at

Bars run from 50 — the average of the models each source lists — so right of the line is above par. The middle column is the gap to whoever leads that scope.

Coding48.6#92/176
Agentic coding48.5#88/169
Agents51.9#94/289
Web research56.8#15/49
Long-horizon work37.9#39/48
Knowledge work60.7#40/198
Computer use44.3#55/81
Reasoning46.1#124/190
Reasoning44.3#91/144
Professional45.2#145/188
Document design31.9#43/48
Core abilities46.2#149/225
Data analysis34.5#41/48
General intelligence51.3#87/214

§ 2 · Head to head

What it beats, and what beats it

The same models turn up scope after scope. Counted once: where both were placed, who finished higher. Bars run right for Muse Spark Open, left for the other.

GPT-4o miniwon 9 · lost 0
GPT-4.1 nanowon 9 · lost 0
Llama 4 Maverick Instructwon 9 · lost 0
NVIDIA-Nemotron-3.5-Lightning-30B-A3B (NVFP4)won 14 · lost 0
Gemma 4 31Bwon 14 · lost 0
Claude Haiku 4.5won 13 · lost 1
NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4)won 12 · lost 2
Nemotron 3.5 Super VLwon 11 · lost 3
GPT-5.4 nanowon 7 · lost 2
Gemini 3.5 Flash-Litewon 7 · lost 2
GPT-4.1 miniwon 7 · lost 2
Mistral Large 3won 7 · lost 2
MiniMax M3won 10 · lost 4
GPT-5.4 miniwon 6 · lost 3
GPT OSS 120Bwon 6 · lost 3
DeepSeek V3won 6 · lost 3
Beam 1.0won 9 · lost 5
Inklingwon 9 · lost 5
Inkling Smallwon 5 · lost 4
DeepSeek V3.2won 5 · lost 4
GPT-5 miniwon 5 · lost 4
Gemini 3.1 Prowon 7 · lost 7
Mistral Large 4won 6 · lost 8
Claude Sonnet 4.6won 5 · lost 9
DeepSeek R1won 3 · lost 6
Gemini 2.5 Prowon 3 · lost 6
Gemini 3.5 Flashwon 3 · lost 6
Claude Sonnet 4.5 (High, 32K, 20250929)won 3 · lost 7
Kimi K2.5won 3 · lost 7
Qwen3.8 27Bwon 4 · lost 10
Muse Spark 1.2won 4 · lost 10
Qwen3.8 Flashwon 3 · lost 10
O3won 2 · lost 7
GPT-5won 2 · lost 7
GPT-5.6 Lunawon 3 · lost 11
DeepSeek V4 Flash Visionwon 3 · lost 11
Gemini 3.8 Flashwon 3 · lost 11
Gemini 3.7 Flashwon 3 · lost 11
GPT-5.5won 3 · lost 11
Claude Sonnet 4won 2 · lost 8
Claude Haiku 5.5won 2 · lost 12
GLM-5.3 Flashwon 2 · lost 12
Step 5won 2 · lost 12
GLM-5.3won 2 · lost 12
Qwen3.8 Maxwon 2 · lost 12
Kimi K3won 2 · lost 12
Grok 4.5won 2 · lost 12
GLM-5.2won 1 · lost 8
DeepSeek V4 Flashwon 1 · lost 8
Claude Opus 4.7won 1 · lost 8
GPT-5.4won 1 · lost 8
GPT-6 Lunawon 1 · lost 13
GPT-5.6 Terrawon 1 · lost 13
MiniMax M3.1 Flashwon 1 · lost 13
Claude Sonnet 5.5won 1 · lost 13
MiMo V2.6 Prowon 1 · lost 13
GPT-6 Solwon 1 · lost 13
DeepSeek V4.1 Flashwon 1 · lost 13
GPT-5.6 Solwon 1 · lost 13
Muse Spark 1.3won 1 · lost 13
DeepSeek-V4-Pro (High, 20260813)won 0 · lost 9
Claude Sonnet 5won 0 · lost 14
Claude Opus 5won 0 · lost 14
Grok 4.6won 0 · lost 14
Claude Opus 4.8won 0 · lost 14
Grok 4.7won 0 · lost 14
Claude Opus 5.5won 0 · lost 14
Muse Wonderwon 0 · lost 14
Claude Fable 5.1won 0 · lost 14
Gemini 4 Argonwon 0 · lost 14
GPT-6.1 Solwon 0 · lost 14
GPT-6 Astrawon 0 · lost 14
Claude Fable 5won 0 · lost 14

§ 3 · Sources

Where the numbers come from

5 publications, 12 figures. Every one links to the page it was read from.

Artificial Analysis ↗1 measureread 2026-10-21
Artificial Analysis Intelligence Index 39
GDPval-AA ↗1 measureread 2026-10-21
GDPval-AA 1401
AA-Briefcase ↗1 measureread 2026-10-21
AA-Briefcase 1279
ARC-AGI-2 ↗1 measureread 2026-10-21
ARC-AGI-2 57.5%
LiveBench ↗8 measuresread 2026-10-21
LiveBench 57.6LiveBench · Agentic 18.9LiveBench · Reasoning 73.7LiveBench · Coding 73.6LiveBench · Browser Use 36LiveBench · Data Analysis 60.2LiveBench · Research 80.3LiveBench · Design 60.3

Badge

PublicAI Index badge for Muse Spark Open[![PublicAI Index](https://publicai.io/model-index/badge?model=muse-spark-open)](https://publicai.io/model-index/m/muse-spark-open)