How much does game localization cost? Build your own estimate

Anyone who answers the question 'how much does localization cost' with a single figure is guessing, because they do not know your game. Two games with the same genre and the same playtime can differ enormously in how much text they carry, how that text is structured, and how often it changes after launch. The only estimate worth trusting is one you build from your own project.

This article walks through how to do that: counting the text you actually have, sorting it by type, deciding what is in scope, and separating the costs that scale with word count from the ones that do not.

Start by counting your words

Before anyone can quote you anything, you need a number: how many words (or characters, for some languages) exist in your game's exportable text. This sounds obvious, but many teams never actually do it — they estimate from memory or from playtime, both of which are unreliable.

The most accurate source is your own text files or database: export every string table, dialogue script, and item description, then run a word count across the export. If your text still lives scattered across code and design documents rather than a single exportable source, extracting it is itself a task worth doing before you go any further, because it also tells you how ready your project is for translation at all.

Separate UI, narrative, and marketing copy

Not all words cost the same to translate, and lumping them into one total hides that. UI strings — menu labels, button text, tooltips — are usually short, repetitive, and constrained by layout space, which makes them fast to translate but demanding to review for fit.

Narrative text — dialogue, quest descriptions, lore, item flavor text — is where most of the word count in story-driven games lives. It needs a translator who can hold character voice and tone consistently across thousands of lines, which is a different skill from translating a settings menu.

Marketing copy — your store page, trailers, social posts — is small in volume but disproportionately visible, since it is often the first and sometimes only text a prospective player reads before deciding to buy. Treating it as an afterthought after the in-game text is a common mistake.

Decide what is actually in scope

A word count is only useful once you have decided what it covers. Scope decisions change the total more than anything else, and they are choices you make, not facts about the language:

• Full text localization versus store-page-only — the cheapest way to gauge demand before committing to the whole game
• Subtitle-only translation for voiced content versus full dubbing — voice work is a separate, much larger undertaking than text
• Which languages you localize into, and whether you launch them together or stagger them over time
• Whether optional or skippable content (side quests, achievements, developer commentary) is included from day one

Costs that scale with word count, and costs that do not

Translation itself scales roughly with your word count — more text means more time, regardless of who does the work. That is the part a word count actually predicts.

A separate set of costs does not scale with word count at all. Preparing your text for export, adjusting UI layouts to fit longer translated strings, sourcing fonts that cover the target language's character set, and setting up the process by which translated files get back into your build are largely fixed engineering work — a text-light game and a text-heavy one may need nearly the same amount of that work done once.

Reviewing the translated build in context — reading it in-game, not just in a spreadsheet — is a cost that scales with the number of screens and states in your game more than with word count, since a short menu label can be just as easy to get wrong as a long line of dialogue.

The recurring cost nobody budgets for

A single word count only estimates the cost of localizing what exists today. Every patch, every seasonal event, every piece of new content adds words that also need translation, in every language you have already committed to. This recurring cost is often larger over a game's lifetime than the initial localization, and it is the one most commonly left out of a first estimate.

If you plan to keep updating your game, the honest way to estimate is not just 'what does today's text cost' but 'what does my expected rate of new text per month cost, for as long as I plan to keep translating.'