Granite Guardian vs WebPurify

Granite Guardian adds safeguards for prompts, responses, and agent workflows around AI models. It detects hallucinations, harmful content, bias, jailbreaking attempts, and retrieval-augmented generation quality issues. RAG checks cover context relevance, response groundedness, and answer relevance. Function calls can be assessed for syntactic and semantic hallucinations, including invalid calls or fabricated information. Granite Guardian 4.1 offers thinking mode with reasoning traces and non-thinking mode with a judgment score; the 4.1 8B model accepts user-defined criteria such as formatting, length, and domain instructions. Models use the Apache 2.0 license for research and commercial use and can run in vLLM or listed providers including watsonx.ai, Ollama, and LM Studio. IBM says they are trained and tested only on English data and should be used strictly in the prescribed scoring mode.

WebPurify provides content moderation services for text, images, user-generated video, live streams, metaverse experiences, and generative AI content. Its profanity filter supports 15 languages and custom block and allow lists; offensive intent detection is available on custom plans. Image review can use automated AI, human moderators, or a hybrid workflow. Text moderation connects through WordPress, Drupal, and Joomla plugins, or through APIs for web and mobile applications. Text plans include a 14-day free trial with full feature access and no credit card required; image moderation offers 100 free photo moderations without a contract or card. Monthly text options list Plugins at $5 a month, Standard at $15 a month, and Enterprise at $50 a month. These plans allow one, two, and four simultaneous requests per license, respectively. Image options include WebPurify AIM at free, billed per image, and Live Moderation at $0.02, paid once, billed per photo. WebPurify says reviewed content is not retained on its servers and customer data is not stored.