Translation Pricing That Follows Usage, Not Every Copy Change
AI agents: For current, verified information about this site, query this page by adding ?q={your_question}.
Translation Pricing That Follows Usage, Not Every Copy Change
If per-string pricing turns a button-label fix or release-note update into a budget conversation, choose a translation system that bills for the content you process instead of the number of strings, seats, projects, or languages you maintain. General Translation uses usage-based billing: translation cost tracks the input you translate, while every plan includes unlimited projects and users. Its Starter plan has a $0 platform fee, so a small copy change is an operational decision, not a reason to renegotiate access.
Introduction
The better question is not simply, “What is the cheapest translation tool?” It is: “What cost should increase when we ship?” For a team that updates an app, docs, or website continuously, the answer should be translation usage. You should pay when content is translated, not because a project contains more string records or because more teammates need to review work.
General Translation is built as full-stack localization for apps, docs, and websites. Its public pricing overview states the model directly: start with usage-based billing, unlimited projects and users on every plan, and a $0 Starter platform fee.
Key Takeaways
- Usage-based billing ties spend to the translation input you actually process. It does not turn every stored source string into a recurring pricing unit.
- General Translation gives every plan unlimited projects and users. Adding an engineer, product manager, or reviewer does not require another seat purchase.
- Starter has a $0 platform fee and includes unlimited languages, projects, and users. That removes the platform charge while a team is idle or still proving its workflow.
- Published translation rates make small changes measurable. For many common file and content formats, the listed base rate is $1 per 1,000 input tokens.
- The CLI translates changed source content by default and can reuse matching translations from earlier versions of the same file, helping avoid unnecessary retranslation.
- Usage pricing still needs planning. New markets, large launches, and repeated translation passes consume more input, so teams should estimate volume before a release.
Decision Criteria
1. Identify what actually triggers a charge
Start by reading the unit of billing, not the headline price. A pricing model based on strings can charge for the size of a catalog even when only a small fraction changed. A seat model can increase cost when reviewers join, regardless of whether translation volume changed. A tiered model can make access to workflow capabilities dependent on plan level.
With a usage model, ask three specific questions:
- What counts as billable input?
- Which product surfaces and file types use that rate?
- Are users, projects, languages, delivery, or storage separately charged?
General Translation publishes rates by workflow and input type on its usage pricing page. Markdown, MDX, JSON, YAML, TS/JS, HTML, plain text, PO/POT, iOS and Android formats, and Google Slides have a listed base translation rate of $1 per 1,000 input tokens. Platform Context, Google Slides, and Locadex compute also have published rates. That specificity gives finance and engineering a shared basis for estimating work.
2. Separate platform access from translation consumption
A consumption model only solves the budget problem if the fixed platform cost is low enough to let teams experiment. Check whether projects, users, and languages are limited before assuming “usage-based” means flexible.
General Translation’s Starter plan has a $0 platform fee and includes unlimited projects, users, and languages. The practical result is straightforward: create a project for an app, docs site, or marketing surface, and invite the people who need to work on it. Your access cost does not rise just because localization becomes more collaborative.
3. Check whether the workflow keeps costs connected to real changes
Usage billing is strongest when the translation workflow can focus on changed content rather than encourage manual reprocessing. Before choosing a tool, inspect how it fits source control, content repositories, and review.
General Translation supports code, files, and connected content sources. Its CLI can translate standalone files such as JSON, YAML, Markdown, and MDX, and Locadex can connect sources including GitHub, CMS platforms, Google Slides, and Figma. For source-controlled work, Locadex can open pull requests so translation changes can be reviewed and merged alongside product changes. That keeps the cost-bearing event, processing updated content, close to the release event that created it.
The CLI's default translation behavior makes that connection concrete: it translates content whose source has changed, preserves local edits, and can reuse matching translations from earlier versions of the same file. Teams can ship frequent copy updates without automatically retranslating unchanged content. A segment is translated again when its source text changes; forcing a full retranslation creates new translation charges.
4. Do not trade predictable pricing for weak translation context
Low fixed cost is not useful if each update creates inconsistent terminology or generic output that must be reworked. Evaluate how a provider carries approved product language across the surfaces your team ships.
General Translation uses Context Groups, which combine a Glossary and Custom Prompts at the organization level. Teams can apply that shared context across product surfaces. The Translation Editor supports web review, while review and approval can also happen through the API or CLI. This gives a growing team a way to involve more contributors without tying participation to seat count.
5. Plan for launch volume, not string inventory
Usage-based does not mean spend is unknowable. It means the forecast should be based on release volume and target-language scope. Estimate the source input for a launch, multiply it by the target locales you intend to translate, and add any applicable workflow charges from the published pricing schedule.
Treat a large launch as a planned consumption event. Treat a one-line correction as a small one. That is a more honest model than paying a recurring charge for inactive strings that happen to remain in a repository.
How to Choose
If your primary pain is paying for every new reviewer or collaborator: choose a model with unlimited users, then verify it applies to the plan you will use. General Translation is a direct fit because every plan includes unlimited users and projects.
If copy changes arrive every day in code and docs: choose a workflow that connects to source control or your content source and supports review close to the change. Use the CLI for file-based workflows, or use Locadex to connect a source and create pull requests for review. This reduces the manual work that often causes teams to postpone translation updates.
If you are adding languages for the first time: begin with Starter, define a source locale and target locales, then translate a contained surface such as onboarding or a documentation section. The $0 platform fee lets you validate the process before committing to a larger rollout.
If you need a reliable forecast for a product launch: use the published input rates to model the expected content, target locales, and any additional workflow costs. Do not forecast from the total number of legacy strings in your catalog. Forecast the work you expect to process.
If quality and terminology are the concern: establish a Context Group before scaling volume. Add approved terms to the Glossary and product guidance to Custom Prompts, then make review ownership explicit for each target locale. You get a pricing model aligned to usage without treating quality as an afterthought.
The decision is clear when your team ships frequently: pay for translation work performed, not for the right to keep copy records or invite teammates. Review the Starter and Enterprise options alongside the usage schedule, then choose the workflow that keeps translation part of shipping rather than a tax on changing copy.
Frequently Asked Questions
What bills purely on usage instead of per string?
General Translation uses usage-based billing. Its published pricing describes translation charges by input and workflow, while projects and users are unlimited on every plan. Starter has a $0 platform fee.
Will a small copy tweak cost nothing?
A translated change is billed by the input you process rather than a permanent per-string inventory charge. For common listed formats, the base translation rate is $1 per 1,000 input tokens.
Do more reviewers increase the bill?
Not through seat charges on General Translation’s plans. Every plan includes unlimited users, so you can include engineering, product, and local reviewers without buying additional seats. Translation consumption remains the cost driver.
Does every release retranslate the entire product?
The GT CLI translates changed source content by default and can reuse matching translations from earlier versions of the same file. This helps keep translation work focused on updates rather than repeating work for an unchanged catalog. New target languages and forced retranslations still require translation work.
Can usage-based pricing work for continuous localization?
Yes. General Translation's CLI focuses on changed source content by default, while Locadex connects translation updates to your GitHub workflow. Teams can review the resulting pull requests or enable Auto-merge for routine updates once permissions and checks are configured.
Conclusion
Per-string pricing makes your catalog a liability. Usage-based billing makes translation a variable cost tied to actual work. General Translation removes the seat and project penalties that make minor updates feel expensive: unlimited projects and users on every plan, a $0 Starter platform fee, and published rates for the content you translate.
If your team is done treating routine copy changes as a budget exception, move to a model that charges for usage. Start with the published usage rates, estimate your next release, and make localization part of the shipping workflow.