Every LingoHub subscription plan includes a monthly capacity of AI credits tailored to your team's typical translation volume. However, because localization workloads change, you are never locked into a single fixed tier. Whether you stay within your plan's included limits or increase capacity, LingoHub adapts directly to your requirements while giving you complete visibility over your budget.
Here is everything you need to know about how your AI credits are calculated for AI translation, what drives your ongoing consumption, and how to keep costs entirely predictable.
Credit capacity by subscription plan
Your monthly AI credits limit and capacity depend on your LingoHub subscription:
Platform: Includes 500k AI credits per month (~37.5k words). It also comes with 5 Mio. hosted characters per month (~375k words), 5 Mio. CDN requests per month, and unlimited seats and projects.
Enterprise: Provides custom AI credit capacity, customized hosted characters, and tailored CDN request limits based on your organization's requirements.
For the latest plan details and custom extensions, visit our pricing page. Need additional volume beyond your plan's included allocation? You can easily add extra capacity starting at €6 per month.
Understanding capacity usage
Every time you run an AI translation, you draw from your pool of AI credits. Depending on what you ask LINA to do, which engine you choose, and whether you are processing text or images, a different amount of credits gets deducted from your balance.
Source text
The source text you want to translate forms the basis for calculating your AI credit usage. For example, 1 source character equals 1 credit on an x1 engine. The final credit consumption depends on the multiplier of the selected Linguistic engine.
Context (Glossaries, Style guides, instructions)
To provide tailored translations, LINA takes your supporting materials, such as Glossaries, Style guides, project and segment instructions, and Labels into account.
Free allowance: For every translation request, you get a free allowance. This is either 5,000 characters or 2x your source text volume (whichever number is higher). For most day-to-day translation work, your context stays within this allowance and does not consume additional credits.
Going over the allowance: Any context beyond the free allowance is charged at 25% of its character volume. This means that 1,000 additional context characters count as only 250 characters toward your billable workload before the engine multiplier is applied.
Engine multiplier
Each AI engine has a multiplier based on what the model costs to run. We align these multipliers with our actual engine costs without adding a markup, keeping AI translation pricing transparent.
1 source character equals 1 AI credit on an x1 engine. An x0.5 engine uses half a credit per source character, while an x2 engine uses two credits per source character.
Engine | Credit multiplier |
Anthropic |
|
Claude Haiku 4.5 | x2 |
Claude Opus 4.7 | x14 |
Claude Opus 5 | x14 |
Claude Sonnet 4.6 | x7 |
Claude Sonnet 5 | x6 |
|
|
Google Gemini |
|
Google Gemini 2.5 Flash | x1 |
Google Gemini 2.5 Pro | x3 |
Google Gemini 3.5 Flash Lite | x1 |
Google Gemini 3.6 Flash | x3 |
|
|
Mistral |
|
Mistral Large 3 | x1 |
Mistral Medium 3.5 | x3 |
Mistral Small 4 | x0.5 |
|
|
OpenAI GPT-5 |
|
OpenAI GPT 5.4 | x6 |
OpenAI GPT 5.4 Mini | x2 |
OpenAI GPT 5.4 Nano | x0.5 |
OpenAI GPT 5.6 Luna | x0.5 |
OpenAI GPT 5.6 Sol | x9 |
OpenAI GPT 5.6 Terra | x5 |
The engine multiplier applies to the total translation workload, including your source text and any billable context beyond the free allowance. For example, translating 1,000 source characters on an x0.5 engine uses 500 AI credits, while the same source text on an x2 engine uses 2,000 AI credits.
Follow these steps to check which models are available and see their exact resource multipliers in LingoHub:
Head to the top navigation bar and click on AI.
From the left sidebar menu under the AI category, select Linguistic engines. Here you can see all available systems by provider. The toggle switch on the right side of each row lets you quickly turn specific engines on or off in your workspace. Only those engines that are turned on will be available for LINA.
Image translation (Quick Translations)
When translating images (such as screenshots, banners, and graphics) in Quick Translations, credit consumption accounts for both text processing and background image editing (removing original text and rendering localized layouts):
Fixed fee per image: Each processed image incurs a fixed fee of 1,000 AI credits to cover image manipulation and object removal.
One-time charge: The fixed image fee is charged once per image, regardless of how many target languages you generate for that graphic.
Text volume charge: In addition to the fixed 1,000 credit image fee, the extracted text characters inside the graphic are charged based on the standard text volume and engine multiplier rule (with a minimum of 100 credits per target language).
Estimate transparency: Pre-translation estimates explicitly break down this cost (e.g., "Includes 1,000 AI credits per image"). The estimate is shown in the UI.
AI translation costs: Examples
The following examples illustrate how different translation scenarios affect your AI credit consumption, from everyday text tasks to image localization and large reference projects.
Example 1: Standard text translation using Google Gemini 2.5 Flash (x1)
You translate 5,000 source characters and provide 9,000 characters of context (Glossary, Style guide, and instructions).
Free allowance: max(5,000, 2 x 5,000) = 10,000 characters
Context is fully covered by the allowance
Billable context: 0 characters
Total workload: 5,000 characters
Engine multiplier: x1
Total consumption: 5,000 AI credits
Example 2: Banner image translation into 3 languages using Google Gemini 2.5 Flash (x1)
You upload 1 promotional banner containing 120 text characters and translate it into 3 target languages.
Fixed image fee: 1,000 AI credits (charged once for the image)
Text volume: 120 characters x 3 languages = 360 characters (x1 multiplier) = 360 AI credits
Total workload:
1,000 + 360
Total consumption: 1,360 AI credits
Example 3: Translation with large reference documents using Google Gemini 2.5 Pro (x3)
You translate 30,000 source characters and provide 500,000 characters of reference material.
Free allowance: max(5,000, 2 x 30,000) = 60,000 characters
Context above allowance: 500,000 − 60,000 = 440,000 characters
Billable context (25%): 110,000 characters
Total workload before engine multiplier: 30,000 + 110,000 = 140,000 characters
Engine multiplier: x3
Total consumption: 420,000 AI credits
What is always free?
Context inside the allowance: Regular use of Glossaries, Style guides, Labels, project instructions, segment instructions, and Translation memory (TM) does not consume additional credits.
Repeated text: If you translate the exact same text again (a cache hit) or if it perfectly matches a string already in your Translation memory, it skips credit usage completely.
Bulk aggregation: Every text translation request must have at least 100 characters to cover the baseline cost of an AI call. Bulk jobs are measured on their grand total, so individual minimums never affect them.
Additional AI credits
If your localization workload increases and you exceed your plan's included capacity, you can easily buy more credits. An additional package of 100,000 AI credits costs €6. This price remains the same regardless of the engine you choose. The engine multiplier determines how many credits are consumed for your translation workload.
Controlling spending and setting budgets
To help you monitor costs and prevent unexpected overage, LingoHub allows Workspace Owners to set up custom budgets for additional AI credit usage and CDN requests.
Learn how to configure spending caps and automated alerts in this guide on How to set up budgets.
Learn how to track current consumption or modify existing limits in this guide on How to manage budgets.
Need more help?
Couldn't find what you were looking for? We are here to help! Reach out to our team at [email protected], and we will gladly assist you.