Chats
The chat as the central tool: input area, tools in the "+" menu and Quick Actions at a glance.
The Chat is the central tool of the meinGPT platform. You type a question or task, the selected model answers, and with a few clicks in the input area you extend the chat with file uploads, web research, artifacts or saved prompts from the library.
Note
Chats are private. Only you see your own chats. The meinGPT team has no access either, no automatic sharing happens.
Where to go next
Input area
Model selection, the "+" menu for sources and tools, voice input, sharing a chat and comfort features
Quick Actions
Buttons under a response that apply a follow-up prompt with one click — default actions, creating your own, managing them
AI Models
Which models exist, what the EU flag means and which model fits when
Very long chats
When a chat gets very long, meinGPT can automatically summarize older messages so the chat keeps working quickly and reliably — if your workspace admin has enabled this feature (it's off by default). When it's active, it can feel like the model "forgot" earlier details - from that point on, it works from a compressed summary rather than the full original text. If the earlier part of the chat included an image or a file, that attachment is no longer available to the model afterward, even though it's still visible in the chat history on screen. Because this also shrinks the context window, the in-chat cost warning can stop appearing too, even if the chat has already cost a noticeable number of credits overall - the warning estimates the cost of your next message from the current context size, not the sum of credits already spent.
Frequent questions
Does meinGPT see my chat content?
No, content is private and visible only to you. The meinGPT team has no access to individual chats. For model requests, your inputs are sent to the respective model provider (OpenAI, Anthropic, etc.) according to the contractually agreed data-protection terms.
Why do some models have an EU flag next to them?
The EU flag indicates the model is hosted in an EU region and is therefore GDPR-compliant. If data protection is a priority, prefer models with an EU flag.
Why does a request sometimes cost more than usual?
Reasoning models (e.g. with "Thinking" in the name) consume significantly more tokens per answer than standard models because they run several internal reasoning steps. Long requests or many files in the context also grow the token consumption.
What happens if I switch the model mid-chat?
The model switch only affects the next answer. The previous chat history is passed to the new model as context, so it knows what's going on. The behavior and tone may shift slightly depending on the model.
Quick Actions don't show up, what can I do?
Check three things: (1) is the Quick Actions feature enabled in the organization, (2) do you have at least one active Quick Action, (3) do you have the necessary permissions? If everything checks out, a page reload usually helps.