Frequently asked questions

An AI environment where all the big models come together in one chat. Claude, GPT, Gemini, Llama, Mistral, Grok plus the open-source frontier. Plus agents, image generators, file analysis and an OpenAI-compatible API. One account, one credit.

No. The free account works without a payment method. Only when you upgrade to Pro, Ultra or buy a top-up do you add a payment method. Credit card, Apple Pay, Google Pay, SEPA bank transfer or cryptocurrency.

Create an account, sign in, open the app. Pick a model, send a prompt. No more setup needed. If you want more, define your own agents or create an API key for code applications.

Current versions of Chrome, Firefox, Safari and Edge. Mobile Safari and Chrome on iOS and Android also work. A native app is not required.

The web app is installable as a PWA. On the home screen it behaves like a native app, with offline cache for history.

Yes. Define a model, instructions and tools like web search once, then reuse the agent whenever you need it. You can share agents or export them as a template.

Claude (Anthropic), GPT (OpenAI), Gemini (Google), Grok (xAI), Mistral, plus the leading open-source models like Llama and DeepSeek. Image generators like DALL-E, Imagen and Stable Diffusion. New models appear on launch day.

Yes. In the model dropdown you pick a new model, the previous history is passed along. The new model sees the full context and responds to it.

For everyday questions a fast, efficient model is enough. For code, analysis and long texts a larger reasoning model fits best, for creative work a high-end model. In the model picker you see speed, context size and price per model, so you can choose per task.

On the Free plan smaller models are available. Pro and Ultra unlock every model. Token limits come from your credit, not from arbitrary quotas.

EU-resident hosting is offered by, among others, Mistral and selected Azure-OpenAI deployments. Account data and chat history live in the EU regardless.

Yes. Describe what you need and pick an image model like DALL-E, Imagen or Stable Diffusion. The image appears right in the chat and you can refine it with follow-up prompts.

PDFs, Word and Excel documents, images, screenshots and source code. Drag them into the chat and ask. Tables are read as data, code keeps its structure. No separate pipeline needed.

Pro and Ultra renew a fixed credit each month. Top-ups are one-time and instantly available. Both sit on the same account, both sources are spent in parallel. Optionally plan credit first then top-up, or vice versa.

Depends on model and length. A short question to a small model costs a fraction of a cent. A 200-page PDF analysis with the largest model is in the low single-digit euro range. In the dashboard you see every request with its exact price, split by input and output tokens.

Plan credit rolls into the next month as long as your plan is active. Top-up credit never expires. It also stays when pausing or when switching between plan and top-up.

Yes. Pro and Ultra cancel monthly, no minimum term. The cancellation takes effect at the end of the paid period. Until then you keep full access. Remaining credit stays with you even after cancellation.

Any time. Upgrades take effect immediately, the new plan is billed prorated for the remaining month. Downgrades take effect at the next billing period. Until then you keep the higher plan.

For every purchase a tax-compliant invoice as PDF in the dashboard. With company address and VAT ID if provided. EU reverse charge for B2B customers with valid VAT ID automatically.

On Pro and Ultra your credit renews automatically at the start of each month. If you spend the monthly amount earlier, further requests pause until renewal, unless you refill via top-up or raise your limit.

Same format. Drop-in compatible with the OpenAI SDKs. You only swap base URL and key. The advantage: you reach every model (also Claude, Gemini, Mistral, Grok) through the same schema and share credit with your chat account.

In the dashboard under API keys. One key per project with its own daily or monthly limit. Keys can be deactivated or rotated any time without touching the code.

Yes. SSE streaming, function calling and structured outputs run across every provider in the unified format. Models without native tool support are routed to a fallback.

You can set daily and monthly limits per key. There's also a technical burst limit that protects individual applications from spikes. In the dashboard you see usage and remaining credit in real time.

Metadata (model, tokens, duration, status) is stored for billing and diagnosis. Prompt and response content only end up in your account log and are accessible to you.

Account data and chat history live on servers in the EU. Model requests go to the respective providers (OpenAI, Anthropic, Google, Mistral, xAI), with US or EU hosting depending on the provider.

No. We don't train models on your data. Your content stays your content.

By default as long as your account exists, so you can search the history. You can delete individual chats or the entire history any time.

TLS 1.3 end-to-end. Data at rest encrypted with AES-256. Backend access goes through 2FA-protected admin accounts. Detailed setup in the privacy policy.

Yes. Login via Google, Microsoft or GitHub. Classic email and password sign-in works too. Multiple methods can be linked per account.

In the dashboard under Security. TOTP apps like Authy, 1Password or Google Authenticator are supported. Backup codes are issued on activation for printing or storage in your password manager.

On the login page "Forgot password" is linked. Enter your email, the reset link arrives within minutes. If you sign in via Google or Microsoft, there's no password to reset.

In the dashboard under Account. Deletion takes effect immediately. Account, chats, agents and API keys are removed. Invoice data remains in a protected archive due to legal retention periods. Remaining credit expires with deletion.

One account is intended per person. Whoever wants to separate private and work uses two email addresses. Teams and companies get separate workspaces with single sign-on through the Enterprise area.