How Krixs AI processes prompts, uses model providers, generates content, reports usage, and applies safety controls.
Last updated: June 4, 2026
This notice explains how the artificial intelligence features available through Krixs AI by Krix Studio work, what information is processed, which third-party providers may be involved, and what responsibilities apply when you use generated content.
This notice supplements the Privacy Policy and Terms of Use. If this notice conflicts with the Privacy Policy regarding personal data, the Privacy Policy applies.
Krixs AI may provide features including:
Features and available models may change as providers, product capabilities, and safety requirements evolve.
When you submit an AI request, Krixs AI may process and send the following information:
Requests are currently sent through the TomDaCat AI API, which may route them to the model you select or to a model selected automatically. Depending on availability and your selection, supported providers may include OpenAI, Google, Anthropic, DeepSeek, and other model providers.
The generated response is returned to Krixs AI and may be displayed, saved to chat history, synchronized through Supabase, and included in usage statistics according to your settings.
You may be able to select a particular model, media model, reasoning level, or response mode. When Auto is selected, Krixs AI may choose a mode or model based on the request, feature availability, performance, safety, and provider limitations.
Different models have different capabilities, context limits, prices, credit costs, safety systems, and privacy terms. A model may be replaced, temporarily unavailable, or unable to complete a request.
Krix Studio does not use private prompts, saved chats, uploaded content, or account data to train public AI models without your explicit consent.
If you enable Improve models with my chats, eligible content may be used by Krix Studio to evaluate and improve product quality, safety, or performance. You can change this preference in Settings.
Third-party AI providers process submitted content under their own API, enterprise, or service terms. Their handling of content may vary by provider, account configuration, model, and connection method.
If chat history is enabled, prompts and responses may be saved locally and synchronized to your account using Supabase. You can delete saved chats or disable chat history in Settings.
Temporary chat data is handled according to the notice shown in the dashboard and is not intended to appear in normal saved chat history.
If you connect TomDaCat AI or another provider:
Do not paste credentials, passwords, or secrets into ordinary AI prompts.
Where returned by the provider, Krixs AI may display or store:
Some providers do not return every usage field. Token counts and costs shown by the product may therefore be estimated, incomplete, delayed, or unavailable. Provider billing records and limits take priority over dashboard estimates.
AI-generated content is probabilistic and may be inaccurate, incomplete, outdated, biased, unsafe, or unsuitable for your purpose. Generated code may contain bugs or security vulnerabilities. Generated media may resemble existing works or include unexpected content.
You are responsible for reviewing outputs before using, publishing, executing, or relying on them. AI output does not replace professional legal, medical, financial, security, compliance, or other specialist advice.
Krix Studio does not guarantee that generated content is unique, error-free, lawful in every jurisdiction, or free from third-party rights.
When using AI Services, you must:
You must not use AI Services to facilitate illegal activity, abuse, fraud, malware, phishing, unauthorized surveillance, harmful manipulation, evasion of safeguards, or attempts to bypass account, credit, or access limits.
Krix Studio and its providers may use automated and human-supported safety measures to detect abuse, protect infrastructure, enforce restrictions, and comply with law. These measures may consider account identifiers, prompts, request patterns, IP addresses, errors, and other security signals.
Authorized administrators may restrict AI access for an account, an IP address, or both. A restriction may be temporary or permanent and may include a reason shown to the user. Contact support if you believe a restriction was applied incorrectly.
We may refuse, limit, interrupt, or remove access to a request or feature where reasonably necessary for safety, security, provider compliance, legal compliance, or service availability.
Authorized personnel may review limited AI interactions, activity records, errors, and diagnostic data where reasonably necessary to:
Access is intended to be limited to people who need it for the relevant purpose.
Krixs AI is designed as an assistance tool. It is not intended to make final decisions that produce legal or similarly significant effects for individuals without appropriate human review.
If you use AI output in a workflow that affects another person, you are responsible for adding appropriate human oversight, verification, notices, and appeal processes.
AI Services depend on third-party providers and may be unavailable, delayed, rate-limited, changed, or discontinued without notice. Available models, prices, credits, context limits, and output quality may change.
We may update this notice when AI features, providers, safety controls, or legal requirements change.
For questions about AI Services, data processing, or an access restriction:
Krix Studio
Email: support@krixishere.org
