What the Privacy Badge Means on Krater.ai: Which AI Models Keep Your Data Private
Learn what the Privacy badge means on Krater.ai, which AI providers never train on your data, and how to filter for private models.
The Privacy badge on Krater.ai indicates whether an AI model's provider uses your prompts and outputs to train future versions of their model. A green Privacy badge means the provider has committed to not training on user data, making it suitable for sensitive business or personal information. Krater.ai displays these privacy indicators in the model selector so users can filter for privacy-safe models before sending any confidential content - a transparency feature unavailable on most single-provider AI platforms.
What Does the Privacy Badge Mean?
In Krater.ai's model selector, you'll see a shield icon next to certain models with the label Private. This Privacy badge indicates that the AI provider behind the model has confirmed they do not train on data submitted through their API.
In plain terms: when you use a Private model, your conversations, prompts, and inputs are processed and discarded - never used to improve or train the provider's AI systems.
What "Private" Means
The Private designation is based on each provider's published data usage policies. A model earns the Private badge when its provider has explicitly confirmed that:
- API inputs and outputs are not used for model training
- Your data is processed only to fulfill your request
- Conversation content is not retained beyond what's necessary for the API call
This is a critical distinction. Some AI providers use API data to train future model versions unless you explicitly opt out. Krater.ai's Privacy badge removes that ambiguity by clearly labeling which providers respect your data by default.
Which Providers Are Marked Private
Based on their published API data usage policies, the following providers are currently marked as Private on Krater.ai:
- OpenAI - Does not train on API data (API usage policy)
- Anthropic - Does not use API inputs for training (usage policy)
- Google - Gemini API data is not used for training (API terms)
- Mistral - API data is not used for model improvement (data policy)
- Cohere - Does not train on customer API data (privacy policy)
- xAI - API data is not used for training (terms of service)
This list is maintained and updated as providers change their policies. Krater.ai continuously monitors provider terms to ensure the Privacy badge remains accurate.
How to Filter for Private Models
Finding private models is straightforward:
- Search - Type "private" in the model selector search bar to see all models with the Private badge.
- Filter - Use the privacy filter to show only models from providers that don't train on your data.
- Browse - The shield icon is visible next to every Private model, so you can spot them while browsing normally.
With 350+ models available, many of which carry the Private badge, you'll always have strong options regardless of your privacy requirements.
Why This Matters
Data privacy in AI isn't just a preference - it's often a requirement:
- Enterprise and business use - Companies handling proprietary information need assurance that AI tools won't leak data into training sets.
- Sensitive personal data - When using AI for personal matters - health questions, financial planning, legal queries - privacy is essential.
- Regulatory compliance - Industries subject to GDPR, HIPAA, or SOC 2 requirements need verifiable data handling practices from their AI tools.
- Intellectual property - Writers, developers, and creators want to ensure their original work isn't absorbed into a model's training data.
Krater.ai's Own Privacy Commitment
Beyond the per-model Privacy badge, Krater.ai itself maintains strict data practices:
- We never train on your data - Your conversations are yours. Krater.ai does not use your inputs to train any AI model.
- Delete anytime - You can delete your conversation history at any time from your settings.
- Transparent routing - Krater.ai acts as an intelligent router to AI providers. Your data is passed to the selected model's API and not stored beyond what's needed for your session.
Combined with the per-model Privacy badge, this creates a layered privacy approach: both the platform and the model provider respect your data.
Frequently Asked Questions
What if a model doesn't have the Private badge?Models without the Private badge come from providers whose data usage policies haven't been confirmed to exclude training on API data, or whose policies are ambiguous. This doesn't necessarily mean they train on your data - it means we can't confirm they don't. When privacy is a priority, stick with Private-labeled models.
How often are provider privacy policies reviewed?Krater.ai monitors provider policies on an ongoing basis. When a provider updates their data usage terms, the Privacy badge is updated accordingly across all affected models.
Does Krater.ai store my conversations?Krater.ai stores your conversation history to provide continuity across sessions, but this data is never used for training. You can delete your entire conversation history at any time from your account settings.
Can I use Private models for GDPR-compliant workflows?Private models from providers with strong data handling policies are a good starting point for GDPR-aware workflows. However, full compliance depends on your specific use case, data processing agreements, and organizational requirements. We recommend consulting with your compliance team for definitive guidance.
Choose AI models that respect your data. Browse Private models on Krater.ai →