The model info card shows the cost on a scale of 1 to 5. For the exact breakdown per model, see the cost overview; the price calculator gives you a feeling for your costs.

Credits apply to your entire company; all users share the balance. The rule is: 100 credits = €1. How much a request uses depends on the model and the message length. For reference: an average user sends around 95 messages per month and uses about €3 in credits. All details in the article How do credits work in Beyond Chat?

Your company receives a monthly Base Credit package. Once it is used up, you can buy Flex Credits under Settings → Billing (€25 per purchase, a maximum of 3 purchases per month). Flex Credits never expire. With Auto-Recharge this happens automatically. If the package regularly falls short, increase the package size.

Under Settings → Analytics you can see usage per user and per model. In model management you can disable cost-intensive models or restrict them to specific groups.

Click the three dots in the top right of the assistant, then click Clone. Then adapt the system prompt to your use case or add a knowledge base.

In the settings of the respective asset, open Access rights and share it with a specific group or with your entire company.

Use the Magic Prompt button: it analyzes and improves your input automatically. For system prompts of new assistants, Prompto, the prompt engineer (among the Pre-Built assistants) helps you. You will find in-depth guides in the article Prompts in Beyond Chat.

Switch to a model with a larger context window or start a new chat. You can summarize long chats with the "Memory Mover" prompt and use the result as a briefing for the new chat. Keep in mind: chat history and uploaded documents count toward the context window.

All data you enter or upload in Beyond Chat is stored on secure German servers (Google Cloud). To process your request, it is transferred to the respective LLM server location, processed there and deleted after a defined period. The model info card shows you which models are EU-hosted.

No. Your input is never used to train the AI models.

Beyond the Loop is GDPR-compliant and ISO 27001 certified. You will find all details about our security practices in our Trust Center. In addition, anonymization automatically protects personal data before it reaches a model.

Under Settings → Analytics you can see engagement, power users, credits and messages per user, usage per model and the most popular assistants. All reports can be exported as CSV. Details in the article Analytics: Understanding AI Usage in Your Company.

No. Administrators only see metadata such as the number of messages sent, never the contents of the chats.

In Company Settings, customize the Security Disclaimer and link your AI policy there. Via group permissions you additionally control who may use which features.

First check the cost overview: it lists all integrated models. If a model is listed but not shown to you, it was probably disabled by your admin in model management or is only available to specific groups. If a model is missing entirely, feel free to ask support.

Models respond at different speeds; reasoning models in particular need extra thinking time. Check the speed (tokens per second) in the model info card and pick a faster model if needed, or let the Smart Router choose.

Right here in the chat: open a new chat with Fin and click "I want to report a problem 🚧". We will take care of it promptly.

Go to Settings → Billing → Manage Subscription. In the billing portal you can change the address and company name for your invoices.

The subscription can be cancelled monthly; with an annual plan you save 10%. Upgrades are done via Settings → Billing → Manage Subscription; amounts already paid for the current month are credited automatically.

Browse the Help Center or write to support@beyondtheloop.ai. For questions directly in Beyond Chat, our support assistant Fin is there for you in the chat.