ProTox 3.0 is a webserver for predicting the toxicity of chemicals. It incorporates molecular similarity and machine learning models to predict 61 toxicity endpoints, including acute toxicity, organ toxicity, clinical toxicity, molecular initiating events, adverse outcomes, and toxicity off-targets. All models are validated on independent datasets and show strong performance. ProTox 3.0 is a freely available computational platform for toxicologists, regulatory agencies, computational chemists, and medicinal chemists. The webserver accepts 2D chemical structures as input and provides toxicological profiles with confidence scores and toxicity radar and network plots. It is accessible via https://tox.charite.de without login requirements. The server includes chemical, molecular target, metabolism, and adverse outcome pathway (AOP) knowledge. It offers 61 models for toxicity prediction, including acute toxicity, organ toxicity, toxicological endpoints, toxicological pathways, molecular initiating events, metabolism, and toxicity targets. The models are based on random forest and deep learning algorithms, with various data sampling methods. The server provides results in multiple formats, including CSV, PDF, and image files. The ProTox 3.0 platform is designed to be user-friendly, allowing users to input compound names or SMILES strings, or draw molecular structures. It supports prediction of hypothetical compounds before synthesis. The server includes models for acute toxicity, organ toxicity, toxicological endpoints, toxicological pathways, molecular initiating events, metabolism, and toxicity targets. The models are validated using 10-fold cross-validation and external validation sets. The performance of the models is evaluated using accuracy, AUC-ROC, sensitivity, specificity, and F1 scores. The server is used for drug and chemical design, supporting green chemistry and toxicology. It helps reduce the need for animal testing and provides efficient and cost-effective toxicity prediction. The ProTox 3.0 webserver is a valuable tool for researchers in toxicology, computational chemistry, and drug discovery. It enables the prediction of chemical toxicity for 61 endpoints, aiding in the development of safer and more environmentally friendly chemicals. The server is freely available and accessible to all users.ProTox 3.0 is a webserver for predicting the toxicity of chemicals. It incorporates molecular similarity and machine learning models to predict 61 toxicity endpoints, including acute toxicity, organ toxicity, clinical toxicity, molecular initiating events, adverse outcomes, and toxicity off-targets. All models are validated on independent datasets and show strong performance. ProTox 3.0 is a freely available computational platform for toxicologists, regulatory agencies, computational chemists, and medicinal chemists. The webserver accepts 2D chemical structures as input and provides toxicological profiles with confidence scores and toxicity radar and network plots. It is accessible via https://tox.charite.de without login requirements. The server includes chemical, molecular target, metabolism, and adverse outcome pathway (AOP) knowledge. It offers 61 models for toxicity prediction, including acute toxicity, organ toxicity, toxicological endpoints, toxicological pathways, molecular initiating events, metabolism, and toxicity targets. The models are based on random forest and deep learning algorithms, with various data sampling methods. The server provides results in multiple formats, including CSV, PDF, and image files. The ProTox 3.0 platform is designed to be user-friendly, allowing users to input compound names or SMILES strings, or draw molecular structures. It supports prediction of hypothetical compounds before synthesis. The server includes models for acute toxicity, organ toxicity, toxicological endpoints, toxicological pathways, molecular initiating events, metabolism, and toxicity targets. The models are validated using 10-fold cross-validation and external validation sets. The performance of the models is evaluated using accuracy, AUC-ROC, sensitivity, specificity, and F1 scores. The server is used for drug and chemical design, supporting green chemistry and toxicology. It helps reduce the need for animal testing and provides efficient and cost-effective toxicity prediction. The ProTox 3.0 webserver is a valuable tool for researchers in toxicology, computational chemistry, and drug discovery. It enables the prediction of chemical toxicity for 61 endpoints, aiding in the development of safer and more environmentally friendly chemicals. The server is freely available and accessible to all users.