Detoxify vs Google Cloud Vision API

Detoxify is a free, self-hosted library that provides trained models and code to classify toxic comments. Its original, unbiased, and multilingual models are based on BERT, RoBERTa, and XLM-RoBERTa, respectively. Depending on the model, outputs include categories such as threats, obscenity, insults, identity attacks, and explicit sexual content. The multilingual model is intended for English, French, Spanish, Italian, Portuguese, Turkish, and Russian. The project describes research, fine-tuning on carefully constructed representative datasets, and helping moderators flag harmful content faster as intended uses. It can be installed with pip and called from Python to predict labels for one string or a list. Inference uses CPU by default and accepts a PyTorch device such as CUDA. The project warns that profanity-associated words may be labeled toxic regardless of tone or intent, potentially biasing results against vulnerable minority groups. Inference uses Transformers and PyTorch Lightning; training also requires the Kaggle API to download data. API access can be requested by email. The library is distributed under Apache-2.0.

Google Cloud Vision API provides image-analysis features that developers can integrate into applications through REST or RPC. Its capabilities include image labeling, face and landmark detection, text recognition, object localization, web detection, and explicit-content tagging. Text Detection reads text in images, while Document Text Detection supports dense text, handwriting, and PDF or TIFF files. Object Localization returns labels and bounding boxes for multiple recognized objects. Web Detection can return related web entities, matching image URLs and pages, visually similar images, and a best-guess label. Face Detection locates faces and facial landmarks and supplies likelihood ratings, but it does not recognize specific individuals. Documented workflows connect it with Cloud Storage, Cloud Functions, Pub/Sub, and Cloud Translation API. For Android and iOS development using Google Cloud Vision services, Google recommends ML Kit for Firebase. Limits include 20 MB per image, 10 MB per JSON request, 16 images per synchronous annotate request, and 2,000 images per asynchronous batch request. The first 1,000 units per feature each month are free; subsequent charges vary by feature and usage tier. The listed plan price is free, with Label Detection listed at 1.50 USD per 1,000 units for 1,001–5,000,000 units.