Myra AI Workspace

Myra AI Workspace is an AI platform from Myra Security GmbH in Munich aimed at organisations with sensitive data. An AI gateway sits between employees or applications and the model providers and enforces rules on users, costs and protected content. It also includes a chat and project interface in the browser and as an app for Android and iOS. Available options are open-weight models on Myra infrastructure, models from more than 20 external providers and self-hosted models. Myra describes itself as a German company without a parent company.

That depends on the platform and the model. According to Myra, the EU-governed platform (Myra Cloud) runs on Myra-owned, certified infrastructure in the EU with locally operated models, while for the EU-located platform Myra only states hosting in Europe without naming the operator. External frontier models, including OpenAI, Anthropic, Google, xAI, Microsoft Azure OpenAI and AWS Bedrock as well as DeepSeek from China, process requests at the provider's location, which according to Myra may be outside the EU. Myra makes its statement of no CLOUD Act exposure explicitly only for Myra-hosted models. According to a 2019 announcement, the separately listed Myra CDN platform also has locations in the USA. The product website includes Google Tag Manager.

A colour code in the chat shows which mode a request runs in. Green stands for open-weight models on Myra's own infrastructure with no third-party access. Orange stands for frontier models with automatic masking of sensitive content, and red for frontier models without masking, which can only be used after approval by administrators. Masked items include names, email addresses, phone numbers and IBANs before the request reaches the provider. With bring-your-own-key, billing runs through the existing contract with the respective provider, and the keys are stored encrypted with AES-256.