---
title: "Concept"
description: "How usage-based billing works"
canonical_url: "https://meingpt.com/en/docs/admin-guide/concept"
language: en
---

# Concept

Our goal at meinGPT is to maximize customer value through regular, extensive platform usage. To achieve this, we use a pricing model where you pay a fixed platform base fee plus the actual usage of AI services. This way we have no incentive to artificially keep usage low.

## How credits work

Prices may change over time due to new AI services or vendor price changes.

**One credit equals one euro cent.**
So 10,000 credits equal 100 EUR.

Credits are calculated as precisely as possible without artificial rounding.

## What does an action cost?

Credits are charged based on usage. The cost of an individual action depends on:

- **Model choice** — reasoning models and premium LLMs consume more credits per token than cheaper standard models. You can find the full per-token table on [All models](/en/docs/admin-guide/pricing).
- **Input and output length** — measured in tokens. Long conversations and verbose answers consume more.
- **Additional features** — image generation, translation, audio transcription, workflow steps. Each has its own price (see the pricing page). Tool calls (e.g. search, reading a website) don't have a price of their own — they show up as the extra tokens they add to text generation.

### Example 1: A single chat

A chat with GPT-4o, short history:

| Component  | Calculation                 | Cost         | Credits  |
|------------|-------------------------------|-------------:|---------:|
| Input      | 523 tokens × 2.96 € / 1M      |    0.0015 €  |     0.15 |
| Response   | 1,247 tokens × 11.82 € / 1M   |    0.0147 €  |     1.47 |
| **Total**  |                                |  **0.016 €** | **1.62** |

Unit reminder: 1.62 credits is 1.62 cents — not 1.62 €.

If the AI uses tools along the way (e.g. a search or reading a website), there's no separate flat fee for that — the extra tokens a tool call produces simply count toward the input/output length of the text generation.

### Example 2: A productive workday

A heavy user who actively reaches for premium models and image generation:

| Activity                                                        | Cost       | Credits  |
|------------------------------------------------------------------|-----------:|---------:|
| 30 longer chats with Claude Opus 4.7 (~8,000 tokens per chat)    |    5.54 €  |      554 |
| 5 images with GPT Image 1                                         |    1.50 €  |      150 |
| 2 workflow wizards                                                |    0.20 €  |       20 |
| **Total**                                                         | **7.24 €** |  **724** |

A single chat costs a few cents. Consumption becomes order-of-magnitude relevant only with frequent use of premium models, high-resolution image generation, and complex workflows.

**Want to estimate ahead?** Use the [cost calculator](/en/docs/admin-guide/cost-calculator) before a workshop or rollout to gauge the budget.

## Cents vs. euros — good to know

Across the platform you see amounts almost everywhere directly in **euros** — balance, consumption, and limits are all shown as euro values. Credits are the internal billing unit (one credit = one cent) and only still appear explicitly in a few places, such as the auto-topup threshold.

If you do come across a credit figure: credits are cents. 8,500 credits = 85.00 €. If a number looks surprisingly large, check the unit.

## Related pages

### [Manage billing](/en/docs/admin-guide/billing)

Buy credits, set up auto-topup, maintain billing details, export top-up history.

### [All models](/en/docs/admin-guide/pricing)

Per-token prices for every available LLM, image model, and translation/audio service.

### [Cost calculator](/en/docs/admin-guide/cost-calculator)

Simulate likely monthly costs based on expected usage.

### [Usage limits](/en/docs/admin-guide/usage-limits)

Set per-user-group cost ceilings and manage upgrade requests.