A translation rate is how translators and language service providers price their work. It is the unit cost agreed upon before a project starts, and it determines the total cost once applied to the volume of content being translated. Understanding how rates work helps localization managers budget accurately, evaluate vendor proposals, and make informed decisions about which service level a given content type actually requires.
💰 Common pricing models
Per word is the most widely used model in the industry. The rate is applied to the source word count, the original text being translated, rather than the target, since source length is fixed and predictable. This model works well for large volumes of text where word count is easy to measure. For East Asian languages like Chinese, Japanese, or Korean, this is typically billed as a “per character” rate.
Per hour is used for editing, proofreading, transcreation, consultation, or projects where the scope is hard to quantify upfront. It is common for short, complex, or loosely defined tasks.
Per page appears most often in certified or sworn translation contexts, where documents are measured in standardized pages rather than words.
Flat project rate applies to fixed-scope work where the total effort is agreed upon in advance, regardless of word count or hours. Many providers also apply a “minimum project fee” to cover the administrative overhead of very small tasks (e.g., a single UI string).
🔍 What affects translation rates
Several factors push rates up or down:
- Language pair. Combinations involving widely spoken languages and large translator pools, such as English and Spanish, tend to cost less than rare or specialized pairs with fewer qualified translators available.
- Content type and complexity. General consumer content translates at lower rates than technical, legal, or medical text, which requires subject-matter expertise and greater precision.
- Service level. A full TEP workflow (translation, editing, proofreading) costs significantly more than translation alone or machine translation post-editing. The rate reflects how many specialists touch the content.
- Turnaround time. Urgent projects typically carry a rush premium.
- Translation memory leverage. When a project contains significant repetition or matches existing TM segments, many providers apply discounts using a “weighted word count” grid. This means you pay a fraction of the full rate for matches (e.g., 25% for a 100% match), which is a key reason why building and maintaining a translation memory reduces localization costs over time.
- Translator expertise. Specialists with domain credentials or rare language combinations command higher rates than generalists.
📊 Key points about translation rates
- Rates vary significantly across markets, language pairs, and providers. A rate that seems high for one language pair may be standard or even low for another.
- Per-word rates apply to source content, not target. Target text in some languages can be significantly longer or shorter than the source, which is why billing on source words is the industry standard.
- Machine translation post-editing is typically priced lower than human translation from scratch, reflecting the reduced effort, but the rate depends heavily on MT output quality and the editing workload it creates.
- TM discounts are a meaningful cost lever for teams with recurring or overlapping content. Negotiating weighted rate terms with a TSP before a project starts can reduce costs substantially on large or repeat projects.
- Rates do not always reflect quality. Vetting translators and TSPs through samples, references, and ISO certification is more reliable than using price as a quality proxy.
How translation costs factor into localization planning
For software product teams, translation rate awareness helps with budget forecasting as the product grows. More languages, more strings, and more frequent releases all multiply the rate by a larger volume. This is one of the reasons automation through translation memory, pretranslation, and MT, has such a direct impact on localization economics: it reduces the volume of content billed at full human translation rates.