Drug Target Benchmarking

Benchmark your gene lists against AI model predictions for disease-specific drug target discovery.

TargetBench v2.0

New

Updated benchmark with the most advanced frontier LLMs across 100 diseases, 10 disease groups, and 10 metrics.

  • Includes all core functionalities of v1.0
  • Most advanced LLMs benchmarked: Including GPT-5.4, Claude Opus 4.6, Grok 4.1, Kimi K2.5
  • 10 disease groups supported: Oncology, Neurology, Immunology, Metabolic, Fibrotic, Cardiovascular, Mental & Behaviour, Ophthalmology, Reproductive, and Rare Diseases
  • 100 indications (up from 38 in v1.0), aligned to EFO standard classification
  • 3 new metrics: Average Precision at Top K, Mechanism of Action (MoA) Clarity, Commercial Tractability Score
Open Latest Version

TargetBench v1.0

The benchmarking framework from Insilico Medicine's landmark publication. Benchmark your gene list against TargetPro and leading LLMs as reported in the study.

  • Includes TargetPro — 71.6% clinical target retrieval, 2–3× above LLMs
  • GPT-4o, Grok 3, DeepSeek-R1, Claude Opus 4, BioGPT, Open Targets
  • 38 diseases across 5 groups: Oncology, Neurology, Immunology, Fibrotic & Metabolic
  • Upload gene list in .csv file to benchmark your gene list
Open v1.0