SprayBench

Expanded results from "Stephen Price, Bryer Sousa, Ronan Flynn, Kiran Judd, Elke A. Rundensteiner, and Danielle L. Cote. SprayBench: Benchmarking LLMs on Open-Ended Materials Reasoning in Additive Manufacturing. NeurIPS 2026 AI4Mat Workshop," research conducted while at Worcester Polytechnic Institute.

Benchmark
525 questions
Hardware
DGX Spark
Models
-- configs
RAG
-- settings

Overall Pass Rate

Variants

Open Weight Models

API Models

MoE

Verbosity

Token Generation Speed

RAG Gains

Leaderboard

Primary Results

Primary SprayBench model leaderboard
Rank

Retrieval

RAG

5,205 articles / 80,161 chunks

SprayBench RAG model comparison
Rank

Efficiency

Architecture

Atomix Research Systems

Need a benchmark, evaluator, or materials AI dashboard?

Contact Atomix