‹ PublicAI Index
The LLM benchmark aggregator.
Muse Spark 1.2
Meta
Strongest in Finance (#2 of 162), weakest in Mathematics (#91 of 152). Above par in 18 of 20 scopes. Among the models it meets almost everywhere, it finishes behind Claude Fable 5 and Gemini 4 Argon and ahead of Gemini 3.1 Pro and Claude Haiku 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.
Professional57.7−4.7#11/1941/2
Finance62.1−0.3#2/1621/1
Medical59.5−7#12/1491/1
Legal57.2−8.7#23/1731/1
Cybersecurity54.1−9.8#24/471/1
Biology research42.9−27.8#24/311/1
Human preference65.6−3.2#11/3541/1
Human preference65.6−3.2#11/3541/1
Coding55.7−10.5#29/1792/6
Agentic coding56−10.9#30/1732/5
Agents57−9.4#37/2982/6
Knowledge work61.1−9.6#40/2082/2
Knowledge53.7−8.8#49/1971/3
Academic knowledge58.8−5.1#20/1231/1
Core abilities54.3−11#55/1991/4
Instruction following61.2−12.6#11/621/1
Language50.7−15.9#35/621/1
Reasoning52.7−12.7#70/1953/5
Reasoning59.4−9.6#22/1471/3
Mathematics49−14.8#91/1522/2
§ 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 1.2, left for the other.
§ 3 · Sources
Where the numbers come from
6 publications, 27 figures. Every one links to the page it was read from.
LMArena Text 1493
GDPval-AA 1480
AA-Briefcase 1324
LiveBench 1.0 78LiveBench 1.0 · Reasoning 90LiveBench 1.0 · Coding 77.5LiveBench 1.0 · Agentic Coding 57.6LiveBench 1.0 · Mathematics 91.2LiveBench 1.0 · Data Analysis 76.5LiveBench 1.0 · Language 78.6LiveBench 1.0 · Instruction Following 74.3
SimpleBench 74.5%
Vals · Legal Research Bench 43.75%Vals · LegalBench 85.26%Vals · Finance Agent 60.6%Vals · CorpFin 70.94%Vals · TaxEval 80.38%Vals · MortgageTax 65.42%Vals · MedCode 49.35%Vals · MedScribe 90.06%Vals · BioMysteryBench 64.81%Vals · CyberBench 69.52%Vals · SWE-bench Verified 86.6% (Mini-SWE-agent)Vals · Vibe Code Bench 79.1% (OpenHands)Vals · Code Migration 29.95%Vals · MMLU Pro 88.28%Vals · ProofBench 43%
Badge
[](https://publicai.io/model-index/m/muse-spark-1-2)