🏆 SpectrumLab Leaderboard

A comprehensive benchmark for evaluating large language models on spectroscopic analysis tasks.

📊 Evaluation Levels: Signal Processing, Perception, Semantic Understanding, Generation
🔬 Domains: IR, NMR, UV-Vis, Mass Spectrometry and more
🌟 Multimodal: Support for both text-only and vision-language models

📈 Stats: 18 models evaluated
🏅 Rankings: 🥇🥈🥉 medals for top performers
🔗 Submit: Send evaluation results to contribute your model!
🏷️ Model Type
👁️ Modality
📊 Sort By

🏆 Model Rankings

Select Model for Details

📊 Subcategory Results

🔍 Column Explanations

  • Rank: 🥇 1st place, 🥈 2nd place, 🥉 3rd place, then numbers
  • Type: 🔓 Open Source, 🔒 Proprietary, 📊 Baseline
  • MM: 👁️ Multimodal, 📝 Text-only
  • Overall: Average accuracy across all evaluated levels
  • Signal: Low-level signal processing tasks
  • Perception: Mid-level feature extraction tasks
  • Semantic: High-level understanding tasks
  • Generation: Spectrum generation tasks

📝 Notes

  • "-" indicates the model was not evaluated on that benchmark
  • Rankings are based on overall performance across all evaluated tasks
  • Multimodal models can process both text and spectroscopic images
  • Click on model names and submitters to visit their pages

📊 Task Categories

Signal Level:

  • Spectrum Type Classification (TC)
  • Spectrum Quality Assessment (QE)
  • Basic Feature Extraction (FE)
  • Impurity Peak Detection (ID)

Perception Level:

  • Functional Group Recognition (GR)
  • Elemental Compositional Prediction (EP)
  • Peak Assignment (PA)
  • Basic Property Prediction (PP)

Semantic Level:

  • Molecular Structure Elucidation (SE)
  • Fusing Spectroscopic Modalities (FM)
  • Multimodal Molecular Reasoning (MR)

Generation Level:

  • Forward Problems (FP)
  • Inverse Problems (IP)
  • De Novo Generation (DnG)