AI Translation Quality Evaluation: How to Test and Score AI Models

AI Translation Quality Evaluation: How to Test and Score AI Models

AI Translation Quality Evaluation: How to Test and Score AI Models

As we’ve learned from the previous article of our AI translation know-how series, there is no perfect model to cover any content in a certain niche. To make the right choice, you need to run tests for every project, language pair and even every type of content (like marketing, UI, website).

This approach helps save money — it’s like testing MVP instead of going all out with risks of poor results.

But how should you run tests and evaluate the results? In this article, we’ll share our behind-the-scenes process. You’ll learn:

  • What top models we use for testing

  • How the testing process goes

  • What metrics there are and why they are not equally important

  • Our secret tool to avoid bias in evaluation

  • Why you can’t choose a model once and for all (a real life example)

There are many models. Which ones do you use for tests?

Yes, there are dozens of NMT and LLM models on the market. We stick to the top industry ones and constantly keep an eye on new promising models — recently we’ve added DeepSeek and Grok (x.ai). 

Below are our most popular NMT and LLM models based on the stats for the past 6 months. This infographic includes only those models that were chosen as best-performing for Alconost’s AI translation projects. The higher the percentage is, the more often the model gets selected.

One might wonder: why not just take those four top-performers from this list and only use them? Well, these figures above are valid as of May 2025, but they are not set in stone. In this industry, experimenting and testing is important to catch up with the constant changes. 

For example, the new Grok model brought us a surprise on the testing stage for one of our new projects: it outscored all regular top-performers for such a common language as German! Below are consolidated testing results for this particular project:

Such cases are hard to predict: models evolve, and each project has a unique combination of domain and content type and, of course, client requirements. However, there is one thing repeating our stats: LLM models tend to outperform NMTs.

Testing process step by step

Step 1. 

We carefully review the source content and client requirements and develop a tailored prompt. Then we run a sample text (~1500-2000 characters) out of the client project through all the chosen models into all the target languages.

❗ We compose a sample text out of different parts of the source content to make it representative of the client project. Strings with variables and code should also be included to see how well models can process them. In case the client project consists of several totally different types of content, we’ll test them separately to get accurate results from testing.

Step 2.

The linguists for the required language pairs evaluate each MT version and leave comments and scores. 

Step 3.

Then they post-edit the best version which becomes the reference translation.

The results of testing and linguist evaluation are exported to a spreadsheet. See the image below for a shortened version of the test for Spanish (ES). 

Score refers to the Linguist Evaluation metric and ranges from 10 to 100. 

Step 4.

Once the reference translations are ready, we upload them along with the raw MT versions to a special tool to get the metrics scores. The tool compares the raw MT versions against the post-edited version (reference translation). Here is an example of exported results:

The goal of the metrics is to assess quality (grammar, fluency, mistranslations, usage of tags and placeholders, etc.) and measure the post-editing efforts needed to achieve the desired level of quality. 

The required level of quality is discussed with the client during the pre-testing stage. Non-critical content (help desk materials, internal presentations, etc.) allows more flexibility and may need less post-editing (or even none).

We always take the time to instruct our linguists before they start post-editing. They receive a style guide and post-editing guidelines about what type of errors are critical for this project and what can be ignored.

To sum up, running the tests looks like that ⤵️

Following the evaluation the client receives a comprehensive report:

  • 10+ translation versions from various NMTs and LLMs;

  • A review of each version by a professional native translator, including a score and comments;

  • A list of top-performing AI models with the performance metrics;

  • A final AQI score that consolidates all these results (more on that further down);

  • A list of rates per language based on evaluations results.

What metrics are used for testing?

There are over 100 metrics suggested by researchers, and they all evaluate different aspects of machine translation. Some of them are created to evaluate LLM translations (and not NMT), some focus on fluency assessment, others — on the precision, etc.

Most metrics are reference-based: they are used to evaluate how well machine translation text matches a human reference translation of the same source content.

Reference-free metrics don’t need a reference translation and are used to further cut costs and speed up the process. We also use one of these metrics in our work. However, such metrics are less reliable than reference-based metrics and serve more as guidelines.

To get a comprehensive view of each model’s performance, we at Alconost put together a versatile set of top industry metrics. Combined with a human evaluation score, this approach ensures an objective, biased-free scoring system. 

Reference-based metrics we use:

  • Comet leverages information from both the source input and the reference translation; shows good correlation with human judgement.

  • BERTscore uses a context-aware approach to compare the similarity of the machine translation output and the reference translation.

  • BLEU is based on sentence-to-sentence comparison. It calculates how many n-grams (a sequence of characters or words) in the MT translation match the reference.

  • nTER indicates the level of human effort required for post-editing. It’s crucial for understanding cost and time efficiency in MTPE workflows. 

  • CHrf++ provides useful complementary insights for certain language types (like morphologically complex languages) where word-level metrics fall short.

Reference-free metric we use:

  • Comet QE offers critical error detection and explainable quality estimation. For this metric we use the CometKiwi22 model that is trained on Direct Assessments and MQM annotations and supports sentence- and word-level quality predictions.

COMET QE is important for real-time applications or when references are unavailable. However, it’s biased: it focuses on fluency and may ignore grammar. Sometimes we see that the score by a human translator/post-editor is only 20 while Comet QE gives the score of around 70-80.

Getting over the shortcomings of QE metrics

The problem with quality evaluation metrics is that they have their weaknesses or biases. For example, nTER does not capture fluency, BLEU doesn’t consider grammar and sentence structure, etc.

These metrics were created with different focuses, so they all bear a different ‘weight’ (impact). If these scores are not equal in importance, it’s hard to make a final judgement about the model’s performance. 

We’ve been experimenting with ranging metrics based on their impact and developed our own reliable unified benchmark — AQI. 

AQI (Aggregate Quality Index) is a comprehensive metric that synthesizes various automated and human evaluation methods to deliver an overall assessment of translation quality. AQI ranges from 0 to 100.

Once we’ve received all the scores, we run them through our tool to calculate the final score, AQI. It lets us understand the overall quality of the output at a glance:

AQI gives us a quick and easily accessible interpretation of all metrics along with the linguist evaluation. Most importantly, it’s an objective interpretation.

It happens that models like GPT and Gemini suddenly drop their performance — for example, start processing strings with variables in a weird way. In this case we turn to the test results, and AQI helps us quickly find an alternative model for replacement.

One model to handle all language pairs?

There is a common misconception that you can pick one top-performing model once and for everything. Unfortunately, it’s not that simple, and here is why:

  • A single model can’t handle all language pairs with equal results. Some excel at European languages but fail at Chinese, others can handle Arabic well but struggle with ‘easier’ languages;

  • The combination of language and domain matters. A model may handle Spanish with ease, but struggle when it comes to a combo of Spanish and a specific domain like cryptocurrency or agriculture;

  • Models get constantly updated, and their performance may change. It’s important to keep testing and experimenting to get the best output possible.

A recent example: evaluation of a website blog translation project showed that each language performed best with a different NMT or LLM model. For this project, we chose a mix of models including DeepL, Gemini, ChatGPT, DeepSeek, and Grok. Interestingly, each of the top-performing models achieved the highest scores in at least one language, while all the other models consistently lagged behind.

For projects with many languages, we often have to choose several models to maximize the quality for every language pair. But even then the choice may not be final.

Here is a real life example:

When we implemented MTPE (machine translation post-editing) for Vizor, a popular game developer and publisher, there were 11 languages. Initially all these languages were translated by Alconost’s native-speaking linguists.

We did the tests, showed the results to the client, and they decided to switch to MTPE for 9 languages, keeping two most important ones for native-speaker localization. 

For these 9 languages, two models were chosen: ChatGPT and DeepL. Every 3 months we did new tests to track AI performance and improve the outputs, and here is what happened:

  1. At one point a new version of ChatGPT came out;

  2. Our tests showed that ChatGPT outperformed DeepL, so it replaced DeepL for all languages except Korean;

  3. A new version of Gemini — a new round of tests — and we found out Gemini became the absolute leader in all language pairs (in the project)!

That’s why we do regular testing for all continuous projects. Regular LQA cycles are a must if you want to maintain consistently high quality, but that’s a topic for a different article. We hope our experience was helpful and you now have a better understanding about testing and scoring AI models. See you in the next article of our AI translation know-how series! 


Related Articles

Our Work

See how we help global companies scale their reach.

JetBrains
Software

JetBrains

1,000,000+ words localized into JA, ES, ZH-CN, KO, PT-BR, FR, TR, CS, RU

Microsoft MakeCode
Software

Microsoft MakeCode

Localization of Microsoft MakeCode

TikTok
Mobile Apps

TikTok

100,000+ words localized into NL, FIL, FI, FR, DE, HE, IT, ES, SV, FR-CA, ES-MX for ByteDance

Viber
Mobile Apps

Viber

Localization of Viber messenger

Read case study
GitHub
Software

GitHub

Translation of GitHub guides and materials

Zendesk
Software

Zendesk

Zendesk Knowledge Base localization for multilingual customer support

Read case study
Airalo
Mobile Apps

Airalo

25,000 words localized into AR, ZH-CN, CS, FIL, FR, DE, EL, HE, HI, IT, JA, KO, PL, PT-BR, RU, ES-419, TH, TR, UK

Choco
Mobile Apps

Choco

15 000+ words localized into JA, KO, VI-VN, IT, PL, NL, PT-BR, CS, CA, ZH-CN

Bitrix24
Websites

Bitrix24

100 000 words and counting localized into ES, PT-BR, JA, ZH-CN and 11 more

Harvard University
E-Learning

Harvard University

Localization of online courses for Harvard University

Xsolla
Games

Xsolla

Localization of Xsolla products

Read case study
SafetyCulture
Software

SafetyCulture

5 000–8 000 words per month localized into JA, PL, TH, TA, SV-SE, VI, UK, ID, HI, KO, NO, PT-PT, RO, RU, TR, AR, BN, ZH-CN, ZH-TW, DA, FI, IT, DE, NL, FR, ES-ES, PT-BR, ES-MX

Veriff
Software

Veriff

12 000 words per month localized into ES-MX, ES-419, SO, SI-LK, VI, SL, SK, SR-CS, RO, PT-PT, PL, MS, MK, LT, LV, JA, HI, DE, KA, FR, FIL, NL, ZH-TW, ZH-CN, CA, BG, BN, ES-ES, PT-BR

Bitrix24 / Voice Responses
Media

Bitrix24 / Voice Responses

into DE, EN, ES, PT-BR, RU, UK

Read case study
DocuWare
Software

DocuWare

90,000 words localized into 23 languages for cloud & on-premises document management

Gartic Phone
Games

Gartic Phone

5 000 words localized into JA, AR, TH, CS, ID, FR, DE, ZH-CN, IT, NL, SV, RO, KA, FA, AZ

Bandsintown
Software

Bandsintown

Localization of Bandsintown app

Read case study
Aviasales
Websites

Aviasales

100,000 words localized into 12 languages for flight search platform

Endomondo
Software

Endomondo

10 000 words and counting localized into CS, HI, NO, TR and 12 more

Liferay
Software

Liferay

Localization of Liferay Platform

BattleTech
Games

BattleTech

Localization of the Battletech game

Goat Simulator
Games

Goat Simulator

Localization of the Goat Simulator game

Stellaris
Games

Stellaris

Localization of the Stellaris game

Movavi
Mobile Apps

Movavi

100,000+ words localized into 20+ languages for video editing software

Parimatch
Software

Parimatch

200,000+ words localized into FR, FR-CA, DE, HI, IT, JA, PL, PT, PT-BR, ES, ES-MX, TR for betting platform

Prequel
Mobile Apps

Prequel

Expanded Top-10 photo editing app to 100M+ Gen Z users worldwide

Ultimate Guitar
Mobile Apps

Ultimate Guitar

4,000 words localized into ES with LQA for Muse Group's guitar app

Wildlife Studios
Games

Wildlife Studios

75 000+ localized into FR, DE, IT, KO, RU, TR, PT-BR, ES-MX, UK, RO, AR

App in the Air
Mobile Apps

App in the Air

500,000+ words localized into PT-BR, PT, NL, KO, HI, FR, ES, SV, IT, TR, JA, AR, DE, ZH-CN, ZH-TW

Apptweak
Mobile Apps

Apptweak

100,000 words localized into JA, KO, ZH, FR for ASO analytics platform

Discourse
E-Learning

Discourse

55,000 words localized into ZH-CN, PT-BR, IT, FR, DE, AR, FI, JA, ES for open-source forum platform

Gcore
Software

Gcore

100 000+ words localized into ZH-CN, DE, ES, PT-BR

Grand Hotel Mania
Games

Grand Hotel Mania

100,000+ words localized from RU into 20 languages for hotel simulator game by Deuscraft

IllFonic
Games

IllFonic

IllFonic Inc.

InterSystems
E-Learning

InterSystems

550+ words localized into ES, FR, PT-BR, ZH-CN, JA

Kissflow
E-Learning

Kissflow

140,000+ words localized into IT, TH for low-code/no-code work platform

Klondike
Games

Klondike

50,000 words localized into DE, ES, IT, FR, PL, NL, JA, KO, ZH-CN, ZH-TW, PT-BR for VIZOR APPS

Clue
Software

Clue

Localization of Clue mobile app

Read case study
Dacadoo
Mobile Apps

Dacadoo

100,000+ words localized into 17 languages for digital health platform

My Cafe
Games

My Cafe

400 000 words and counting localized into FR, ES, PT-BR, KO and 6 more

Party Hard
Games

Party Hard

Localization of the Party Hard game

Planner 5D
Mobile Apps

Planner 5D

20,000 words localized into 24 languages for home design app

Punch Club
Games

Punch Club

20 000 words localized into ZH-CN, PL

Read case study
RICOH360 Tours
Software

RICOH360 Tours

18 000 characters localized into Japanese –> English, German, French, Spanish, Dutch

Sumsub
Software

Sumsub

7,000 words localized into 28 languages for identity verification platform

Transporeon
Software

Transporeon

50,000 words localized into 18 languages for logistics visibility platform

Aktiia
Websites

Aktiia

21,000 words localized into FR, DE, IT for blood pressure monitoring startup

Awarefy
Mobile Apps

Awarefy

30 000 characters localized into Japanese –> English

Baby Tracker
Mobile Apps

Baby Tracker

5 000 words localized into ES-LA, PT-BR, DE, UK

Circuit
Mobile Apps

Circuit

5,000 words localized into 30+ languages for delivery route planning app

CSAT
Software

CSAT

200 000+ words localized into AR, HE, IT, KO, PL, PT-BR, PT, TR, ZH-CN

Driivz
Software

Driivz

1 300 words localized into HR, CS, ET, FI, FR, FR-CA, DE, EL, HU, IS, IT, LV, LT, NO, PL, RO, SK, SL, ES-ES, SV

Foodback
E-Learning

Foodback

50,000 words localized into 12 languages for restaurant feedback platform

Gentler Streak
Mobile Apps

Gentler Streak

2 000 words per month localized into FR, DE, IT, ZH, ZH-HK, JA, KO

Harvest Land / Paris: City Adventure
Games

Harvest Land / Paris: City Adventure

200,000+ words localized from RU into 8 languages for Mysterytag games

Harvest Land
Games

Harvest Land

2 000 words per month localized into RU → EN, ES, PT-PT, FR, IT, DE, KO, JA, ZH

Hotel Life
Games

Hotel Life

12,000 words localized into 10 languages for hotel simulation game by Eidolon

HUB Parking
Software

HUB Parking

62,000 words localized into RU for smart parking solutions

Keenetic
Websites

Keenetic

30,000 words localized into PL, ES, FR, DE, SV, PT, IT for Wi-Fi router manufacturer

Charm Farm
Games

Charm Farm

Localization of Charm Farm Game

Read case study
Zombie Castaways
Games

Zombie Castaways

Localization of the Zombie Castaways game

Meisterplan
Software

Meisterplan

74,500 words localized into ES, FR, DE for project portfolio management

Onde
Mobile Apps

Onde

up to 1 000 words per month localized into SV, RW, DA, SQ, PL, KM, ET, MY, ZH-HANS, FI, DE, LV, HE, NL, HR, SK, NO, LT, IT, TH, SO, ID, IS, UR-PK, ZH-HANT, CS, UK and 10 more

OpenProject
Software

OpenProject

1 000+ words per month, up to 150 000 localized into FR, ZH-CN, ES-ES, IT, PL, PT-PT, PT-BR, KO, UK

Pillow
Software

Pillow

100,000+ words localized into 13 languages for sleep tracking app by Neybox

Playwing
Games

Playwing

40 000+ words localized into AF, AR, BN, MY, HR, CS, NL, ET, FR, KA, DE, EL, HU, ID, MS, PL, PT, RU, SK, ES, SV, TH

Clash of Kings
Games

Clash of Kings

Proofreading of in-game text for Clash of Kings

Read case study
Soundiiz
Software

Soundiiz

15,000+ words localized into 14 languages for music playlist transfer app

Speakap
Mobile Apps

Speakap

5,000 words localized into DE, NL, ES for employee communication app

Stripo
Websites

Stripo

25 000 words localized into PT-BR, TR, CS, FR, DE, IT, ES, PL, ZH-TW, NL, SL

Sufio
Mobile Apps

Sufio

3 000 words localized into FR, DE

Tonsser
Mobile Apps

Tonsser

40,000 words localized into ES-US, PT, SV, DE for football community app

Vizor
Games

Vizor

into ES-ES, NL, PL, ZH-CN, ZH-TW, PT-BR, IT, KO, FR, DE

Read case study
Alvadi
E-Commerce

Alvadi

Multilingual SEO for automotive supplier expanding to 30+ markets

BoxHero
Software

BoxHero

10000 localized into ES-419, ZH-CN, ZH-TW

Epic Roller Coasters
Games

Epic Roller Coasters

4,000 words localized into ZH-CN, FR, DE, JA, KO, RU, ES for VR game by B4T Games

Dating Apps Bundle
Mobile Apps

Dating Apps Bundle

50,000 words localized into 36+ languages for Red Panda Labs dating apps

Face Yoga
Mobile Apps

Face Yoga

2,000 words localized into ES-419, PT-BR for skincare app by Tepluhab

Forest Bounty
Games

Forest Bounty

10,000 words localized from RU/EN into ES, FR, PL, PT-BR for VigrGames

Game Localization — MTPE
Games

Game Localization — MTPE

Anonymized case — under NDA

Read case study
HUD App
Software

HUD App

10,000 words localized into 18 languages for dating app

DreamCommerce
E-Commerce

DreamCommerce

Localization of DreamCommerce Platform

Read case study
Jooble
E-Learning

Jooble

10 000 words localized into ES, PT, KO, JA and 11 more

Read case study
Smarty CRM
Software

Smarty CRM

Localization of Smarty CRM platform

Read case study
Targetprocess
Software

Targetprocess

Localization of Targetprocess platform

Mahjong Treasure Quest
Games

Mahjong Treasure Quest

30 000 words localized into EN → JA, MTPE: EN → PL, NL, KO, ZH-CN, ZH-TW, DE, FR

Primagest
Websites

Primagest

80 000 characters localized into JA → EN, ZH

Raymy
Software

Raymy

80 000 characters localized into Japanese –> English, Chinese (traditional),Vietnamese, Hindi

Sana Commerce
E-Commerce

Sana Commerce

Bi-weekly B2B e-commerce platform updates in 22 languages

Swappy Dog
Games

Swappy Dog

25,000 words localized from RU into 19 languages for match-3 game by Funmatica

Swoo
Mobile Apps

Swoo

30,000 words localized into ES, IT, PT for digital wallet app by CARDS/MOBILE

EnjoyGaming
Games

EnjoyGaming

500 words per month localized into DE, ES, FR, HI, IT, JA, KO, PT, PT-BR, RU, SV, TR, UK

2Solar
Software

2Solar

10,500 words localized into DE for solar software platform

24 Hour Home Care
Software

24 Hour Home Care

2,590 words localized into ES-419 for healthcare staffing company

ActiveMap
Software

ActiveMap

18 000 words localized into AR

Adizes Institute
E-Learning

Adizes Institute

5,850 words localized into HE for leadership consulting platform

AI Chat Smith
E-Learning

AI Chat Smith

1 500 words per month localized into ES, JA, RU, ZH, DE, FR, PT-BR

Alice VR
Media

Alice VR

8 phrases localized into CA, EN, ES, RU

Read case study
Appewa
E-Learning

Appewa

100+ words localized into 20 languages for language learning app by Lithium Lab

Associations
Games

Associations

3 000 words localized into TR, PL, SV-SE, NO, DA, CS, SK, HU, JA, KO, and 7 more

Aviloo
Software

Aviloo

5,000 words MTPE from DE into DA, NL, FR, IT, SV, NO for EV battery diagnostics

Berry Factory Tycoon
Games

Berry Factory Tycoon

1 500 words every two months localized into RU → EN, KO, JA

BestChange
Websites

BestChange

2 000 words per month localized into NL, PL, SV

Blink
E-Learning

Blink

32 300 words localized into FR

Bunny Boom
Games

Bunny Boom

3 000 words localized into DE, ES, FR, IT, JA, KO, PT-BR

Life is Feudal
Media

Life is Feudal

Character voiceovers for Life is Feudal: Your Own

Read case study
Conf.app
Software

Conf.app

4,500 words localized into IT, ZH-CN, PT-BR, DE, ES for event management app

Cosmos VR
Media

Cosmos VR

2 000 words localized into CA, DE, EN, ES

Read case study
Darksy Cleaner
Mobile Apps

Darksy Cleaner

1,400 words localized into 9 languages for iOS photo cleaner app

Days After
Games

Days After

500 words every 1.5 months localized into RU → EN, PT-BR, ES; EN → DE, FR, KO, AR, ZH-TW, ZH-CN, NO, PL, TH, CS, JA and 10 more languages on demand

Dople
Software

Dople

11 500 characters with space localized into KO → JA

eSIM Provider
Websites

eSIM Provider

around 30 000 words when requested localized into SQ, AR, HU, IT, IS, NL, FR, DE

EXR
Games

EXR

12 000 words localized into ES, FR

GoodCrypto
Software

GoodCrypto

2 000 words per month localized into AR, ZH, FR, DE, ID, IT, KO, PT-BR, ES, TR, VI

Haiku
Games

Haiku

10 000+ words localized into ES-419, PT-BR, DE, JA, ZH-CN

Impulse
E-Learning

Impulse

Impulse - Brain Training

IQ Dungeon
Games

IQ Dungeon

IQ Dungeon - Riddle Solving RPG

Knights and Brides
Games

Knights and Brides

Knights & Brides

Lexilize
E-Learning

Lexilize

7 000 words localized into FR

Darklings
Games

Darklings

1 000 words localized into JA, ZH, ES, RU, IT, FR, DE, PT, KO

Kill Shot Bravo
Games

Kill Shot Bravo

Localization of Kill Shot Bravo

Next Stop
Games

Next Stop

7 500 words localized into FR, DE, EN, JA

EcoCity
Games

EcoCity

Localization of the EcoCity game

Forced Showdown
Games

Forced Showdown

Localization of the Forced Showdown game

Minion Masters
Games

Minion Masters

Localization of the Minion Masters game

Outpost Zero
Games

Outpost Zero

Localization of the Outpost Zero game

Streets of Rogue
Games

Streets of Rogue

Localization of the Streets of Rogue game

Tamadog
Games

Tamadog

Localization of the Tamadog game

Valentine's Day
Games

Valentine's Day

into DE, FR, IT, ES, PT-BR

Mimic Logic
Games

Mimic Logic

13 000 characters localized into JA → EN, ZH-CN

Mini Golf 100+
Games

Mini Golf 100+

10 000 characters localized into Japanese –> English, German, French, Spanish, Korean, Chinese (tw), Chinese (zh), Portuguese (Brazil)

Mini Mini Farm
Games

Mini Mini Farm

8 500 characters localized into Japanese –> English

mod.io
Games

mod.io

500 words localized into ZH-TW, ZH-CN, DE, IT, JA, KO, PL, RU, ES

MySignature
Websites

MySignature

1 500 words per month localized into IT, FR, NL, FI, PL, DE, ES, PT

Parasite Days
Games

Parasite Days

70 000 characters localized into Japanese –> English

PDIS
Software

PDIS

2 346 characters with space localized into KO → EN

PosterMyWall
E-Learning

PosterMyWall

1 000 words per month localized into ZH-HANS, DA, NL, FR, DE, ID, IT, PL, PT, RU, ES, TH

Prospre
Software

Prospre

7 000 words localized into ZH-CN, FR, DE, IT, JA, PT-BR, ES-419

Ruins Magus
Games

Ruins Magus

38 000 characters localized into Japanese –> English

Samedi Manor
Games

Samedi Manor

2,000 words localized from RU into 7 languages for idle game by Black Caviar Games

Soltec Health
E-Learning

Soltec Health

17 000 words per 6 months localized into JA

Soma Development
Software

Soma Development

8 000 words localized into AR, ZH-CN, FR, DE, ID, IT, JA, PT, RU, VI, ES-419

Sonnet of Wizard
Games

Sonnet of Wizard

224 261 characters localized into Japanese –> English

Sportplus
Websites

Sportplus

800 words localized into AR, HI

Hotel Project
Games

Hotel Project

3,622 words localized into PT-BR for merge game by Next Epic

Tovie AI
Software

Tovie AI

4,800 words localized into ES, PT-BR for conversational AI platform

Ultight
Software

Ultight

5 046 characters with spaces localized into KO → EN

Underground Waifus
Games

Underground Waifus

4 300 words localized into JA, ZH-CN, KO, FR, IT, DE

UNNI
Software

UNNI

15 000 words per month localized into TH

Vlad & Niki
Games

Vlad & Niki

15,000 words localized into 10 languages for kids claymation game by RUD present

Kerish Doctor
Media

Kerish Doctor

Voiceovers for the Kerish Doctor software

Read case study
Welcome Bot
Software

Welcome Bot

2 000 words localized into UK, LT, AR, ES, FR, DE, PT, IT, PL, HE, ID, TR, HI, VI, MS, TH, CS, NL

WRD
Media

WRD

WRD – Learn Words App Voiceover

Read case study
Azur Games
Games

Azur Games

200 – 500 words per order localized into ID, PL, IT, TR, ZH-CN, ZH-TW, KO, PT-BR, JA, FR, ES, DE, TH, HI

Character Bank
Software

Character Bank

Localization for Character Bank software platform

Coffee Break
Software

Coffee Break

Localization for Coffee Break software platform

Google
Software

Google

Localization for Google

GROOVE
Software

GROOVE

Localization for GROOVE X

Hakali
Software

Hakali

Localization for Hakali

Request a Quote

Whether you're launching in new markets or scaling existing localization — let's make it happen.

This field is required
This field is required
Please enter a valid email address
Please enter a valid phone number
This field is required
This field is required