You might also be interested in learning about AI-powered language quality assurance and the AI suggestions features.
AI Translations is a new task type that lets you translate content quickly, accurately, and at scale — with full automation built directly into the Lokalise interface. You can create and manage AI translation tasks right from your project dashboard or via the API, without switching to other tools.
Powered by the AI assistant which uses advanced models like GPT-5 and Claude translations are enriched with your project’s context, glossary, and style guides. With smart routing across multiple LLMs, Lokalise automatically selects the best model for each language pair and content type, ensuring consistently high quality.
What AI Translations offers
Translate content into dozens of languages and variations.
Tasks are automatically assigned to the Lokalise AI assistant, with keys locked until completion.
Add custom instructions (e.g., tone of voice, formality, target audience).
Ensure glossary terms and existing terminology are respected.
Receive notifications as soon as translations are finished.
Translations delivered up to 10× faster than traditional human workflows.
Save up to 80% in costs while maintaining scalability and quality.
Only 20% of output requires post-editing, flagged automatically via quality scoring.
Most tasks take just a few minutes to complete, depending on content volume and number of target languages. You’ll see the estimated completion time once you set the scope.
How AI in Lokalise delivers human-like translations
Just like a professional translator, our AI solutions produce translations that sound natural, not robotic or overly literal. This is achieved through AI orchestration:
Context-aware translation – each task is enriched with your glossary, translation memory, and style guides, so terminology, cultural nuance, and brand voice are preserved.
Smart model routing – content is dynamically routed to the most suitable large language model (e.g., GPT-5, Claude, or others), depending on the language pair and context.
Fluency and accuracy – the result is content that is precise yet reads as if written by a native speaker.
In around 80% of cases, translations are ready to publish immediately, without further edits.
Currently, Lokalise does not support customization or fine-tuning of integrated AI models, such as GPT or Claude.
Supported languages
AI Translations currently supports the following languages and their variations for different locales:
Supported locales
Supported locales
Afrikaans
Albanian
Arabic
Armenian
Assamese
Azerbaijani
Bashkir
Basque
Belarusian
Bengali
Bihari
Bosnian
Bulgarian
Catalan
Cebuano
Corsican
Croatian
Czech
Danish
Dutch
English
Esperanto
Estonian
Faroese
Finnish
French
Frisian
Galician
Georgian
German
Greek
Gujarati
Haitian Creole
Hebrew
Hindi
Hungarian
Icelandic
Indonesian
Interlingua
Interlingue
Ido
Irish
Italian
Japanese
Javanese
Kannada
Kashmiri
Kazakh
Khmer
Kabuverdianu
Konkani
Korean
Kyrgyz
Lao
Latin
Latvian
Limburgish
Lithuanian
Luxembourgish
Macedonian
Malagasy
Malay
Malayalam
Maltese
Mauritian Creole
Mongolian
Marathi
Neapolitan
Nepali
Northern Sami
Norwegian
Occitan
Oriya (Odia)
Pashto
Persian
Piedmontese
Polish
Portuguese
Rhaeto-Romanic
Romanian
Russian
Sardinian
Scottish Gaelic
Serbian
Sindhi
Sinhala
Slovak
Slovenian
Somali
Spanish
Sundanese
Swahili
Swedish
Tagalog
Tamil
Telugu
Thai
Tatar
Turkish
Turkmen
Ukrainian
Urdu
Uzbek
Vietnamese
Welsh
Walloon
Xhosa
Yiddish
Yoruba
Cantonese
Chinese Simplified
Chinese Traditional
Zulu
Note: This is not an exhaustive list. It will be updated based on the quality level and our confidence in each language. If you'd like us to support additional languages and are willing to help with translation quality evaluation, please let us know!
Using AI Translations
Creating a new AI Translations task
Keep in mind that AI Translations consume Pro AI words quota. If the team that the project belongs to does not have any Pro AI words left, you won't be able to create the AI Translations task. Refer to the Team quotas article to learn more.
To get started with AI Translations, open any project in Lokalise and create a new task. There are two ways to create a task.
Via the editor
It's important to remember that a translation key cannot be assigned to a new task if it is already part of another ongoing task. To include such a key in a new task, you have two options:
Wait for the existing task to be completed: Once the task is finished, the key will be available for reassignment in a new task.
Remove the key from the current task: If you need to reassign the key immediately, go to the project editor, select the key, and use the Remove from task option available in the bulk actions menu. This will free up the key, allowing it to be included in a different task. However, if the key is already marked as completed in the task, it won't be possible to remove it from that task.
Start by selecting multiple keys in the editor. You can do this by ticking the checkboxes next to the keys you want to include in the task.
Once selected, choose Create task... from the bulk actions menu. This will take you directly to the task creation page, where the task scope is automatically set to the keys you’ve selected.
Via the Tasks page
Alternatively, you can create a task from the Tasks page in your project. Simply navigate to your project and go to the Tasks page:
Click Create a task. This will open the task creation wizard, where you can define the task's details and scope.
Providing general task information
In the task creation wizard, you'll find a new task type called Automatic translation. Select this type to begin.
Start by giving your task a name and adding task instructions. These instructions provide high-level context for the translation and help our AI produce results that are better aligned with your needs. Think of them as a description of the task, not a detailed set of linguistic rules.
The instructions field has a 2,000-character limit.
Can't I just use task instructions to control tone or terminology?
We don’t recommend it. Task instructions are best for high-level context (e.g., what the content is and where it will be used).
For tone, formality, or terminology, attach a style guide to the task. Our AI will use it as additional context, alongside your project glossary, which is applied during the translation process.
Enterprise customers can also take advantage of custom AI profiles to provide even more context.
Using task instructions like prompts may seem to work in small cases, but it often produces inconsistent results across languages and larger volumes. Our AI performs best when rules are structured, centralized, and applied systematically through style guides and glossaries.
Good uses of task instructions:
“This task contains UI copy for the mobile banking app dashboard.”
“Includes labels and tooltips for e-commerce product pages.”
“Internal help desk content for IT support.”
What task instructions are not:
❌ A replacement for your style guide or glossary
❌ A place to define terminology, spelling, or brand rules
❌ A prompt-style request to the AI
Adjusting advanced task options
When setting up an AI Translation task, the advanced options are limited to the following:
Apply 100% translation memory matches — when enabled, Lokalise will search for exact (100%) matches in the translation memory for all strings included in the task. If a match is found, it is applied automatically. AI will not modify these segments and no AI words are consumed.
Translate all segments (including existing translations, except 100% TM matches) — available only for segmented projects. When unchecked (default), AI translates only empty segments and skips non-empty ones. This is the new default behavior for segmented projects. When checked, AI translates all segments, including those that already have translations.
Note that the Apply 100% translation memory matches option (enabled by default) always takes priority. Even if Translate all segments is enabled, any segment with a 100% TM match will be filled from translation memory and will not be sent to AI, ensuring consistent results and avoiding unnecessary AI word usage.
Mark translations as verified — when enabled, all translations produced by AI during the task will be marked as verified.
Save translations performed by AI to Translation memory — choose whether translations generated by AI should be saved to your translation memory for future reuse.
Tag keys after the task is closed — enable this option to automatically tag all translation keys included in the task once it is completed, helping you track and manage them later.
Adjusting task scope
Select the scope and languages:
Task scope — adjust the filter to select the specific keys that need to be included in the task. This ensures that only the relevant content is translated.
Source language — choose the language that will serve as the reference for the translations. This is the language from which AI will translate.
Target languages — select one or more languages that you want the AI assistant to translate into.
Ensure that you choose the correct locale for each target language. Incorrect or unspecified regions can lead to unexpected translation results. For instance, selecting "Portugal" might cause the AI to use Brazilian Portuguese. To avoid this, make sure to choose "Portuguese (Portugal)" for translations intended for that region.
Task assignees — you cannot modify assignees in AI Translation tasks, as all translations will be automatically handled by the Lokalise AI assistant.
Task summary
On the right side of the task creation wizard, you'll see the task summary along with your Pro AI words balance.
Pro AI words quota
This section shows how many Pro AI words your team has available. Below (in the task summary) you'll also see how many words will be consumed by the task.
Pro AI usage is based on input: the number of source-language words multiplied by the number of target languages included in the task. In addition to Pro AI words, every AI-powered task also consumes processed words, which are counted from the output — the total number of words the model generates across all target languages.
Example:
If your task includes 40 source words and you assign it to 3 target languages, the task will consume:
Pro AI words: 40 × 3 = 120 Pro AI words
Processed words: depends on the output (e.g., 38 + 41 + 40 = 119 processed words)
For more details on how quotas work, refer to the Team quotas article.
Estimated delivery time
The summary also provides an estimated delivery time for the task. This is an approximation and may vary depending on current system load.
Frequently asked questions
💡 Looking for more?
Find all answers in AI: Frequently asked questions!







