70% of translations are machine-assisted
Machine-assisted translation methods are now used for 70% of translations, overtaking human translation which is declining rapidly.
Download full report
AI translation is on the rise as large language models outperform machine translation tools
Download the full trends report for insights, trends, and best practices that will help you and your stakeholders optimize localization processes in 2025.
Claude 3.5 ranks #1 in translation accuracy
Initial research from the product team at Lokalise showed that large language models (LLMs) outperformed machine translations in translation accuracy. Claude 3.5 ranked #1 in all results, significantly outperforming all other language models across all languages that were tested.
Tested via Lokalise AI
| Rank | Engine |
| 1 | Claude 3.5 |
| 2 | Google Translate |
| 3 | GPT |
| 4 | DeepL |
| 5 | Microsoft Translator |
Translation memory is the rising star in localization
Companies are increasingly repurposing translations to reduce costs, maintain quality, and speed up translation processes. It also indicates a clear preference for more automated, scalable translation methods over human translation.
Download full report
There’s growing trust in AI translation
There’s a notable trend in small and mid-size companies adopting AI for translation in 2024.
Download full report
Localization trends report 2025
Discover trends and best practices to help you localize more efficiently this year. Download the report now.
