‹ PublicAI Index
The LLM benchmark aggregator.
Muse Spark 1.3
Meta
Strongest in Instruction following (#4 of 62), weakest in General intelligence (#72 of 180). Above par in 21 of 23 scopes. Among the models it meets almost everywhere, it finishes behind Gemini 4 Carbon and Claude Fable 5.1 and ahead of Claude Opus 4.8 and GPT-5.5.
§ 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.
Human preference65.9−2.9#10/3541/1
Human preference65.9−2.9#10/3541/1
Coding57.7−8.4#21/1794/6
Agentic coding58−8.7#19/1734/5
Core abilities58−7.6#22/2002/4
Instruction following67.7−6.1#4/621/1
Data analysis57.3−8.8#10/321/1
Language56.2−10.4#21/621/1
General intelligence51.8−17.2#72/1801/3
Reasoning57.9−7.5#22/1954/5
Reasoning61.2−7.8#16/1472/3
Mathematics55.5−8.3#46/1522/2
Professional54.4−8#38/1942/2
Legal61−4.9#9/1731/1
Cybersecurity57−6.9#16/471/1
Document design52.4−16#16/321/1
Finance57.8−4.6#20/1621/1
IT operations42−31.8#20/331/1
Agents55.1−11.2#55/2983/6
Long-horizon work55−18.8#10/321/1
Knowledge work66.2−4.5#13/2082/2
Web research52.8−10.8#16/361/2
Computer use39.9−28.2#55/691/2
§ 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 1.3, left for the other.
§ 3 · Sources
Where the numbers come from
8 publications, 28 figures. Every one links to the page it was read from.
LMArena Text 1496
GDPval-AA 1681
AA-Briefcase 1581
Terminal-Bench 14.8% (Muse Code)
LiveBench 68.6LiveBench · Agentic 47.9LiveBench · Reasoning 89.7LiveBench · Coding 84.4LiveBench · Browser Use 30.9LiveBench · Data Analysis 79.6LiveBench · Research 82.5LiveBench · Design 74.8
LiveBench 1.0 81.6LiveBench 1.0 · Reasoning 89.7LiveBench 1.0 · Coding 81.1LiveBench 1.0 · Agentic Coding 64.1LiveBench 1.0 · Mathematics 95.9LiveBench 1.0 · Data Analysis 79.6LiveBench 1.0 · Language 82.8LiveBench 1.0 · Instruction Following 78
SimpleBench 81.8%
Vals · Legal Research Bench 55.29%Vals · Finance Agent 59.96%Vals · CyberBench 72.74%Vals · SRE Bench 3.82%Vals · Vibe Code Bench 85.86% (OpenHands)Vals · Code Migration 47.41%Vals · ProofBench 58%
Badge
[](https://publicai.io/model-index/m/muse-spark-1-3)