Translation on Okou
Text in one language, the same text in another, at a price low enough to translate everything rather than only what seems important.
Generation · No setup · Managed credits per successful call
Give it text and a target language and it gives the text back translated. The source language is worked out for you unless you want to state it, and only the translation comes back, so it drops straight into whatever the agent is building.
It runs on a managed model chosen for speed and cost, which makes it the right tool for volume: every incoming message, every review, every subtitle, rather than the few things somebody flagged. Okou holds the account and charges by the length of the text.
What Translation is
Most translation in a company is not the polished kind. It is understanding what a customer wrote, reading a competitor's page, or making a note legible to the half of the team that does not share a first language.
That work is high volume and low ceremony, and it stops happening when each translation is a task for a person or a paid seat. Making it a cheap call an agent can make means it happens by default.
This service does exactly that: text in, translated text out, source detected, nothing else in the response. It is deliberately plain, because it is usually one step in the middle of something larger.
For copy that carries your brand or your legal exposure, this is the draft rather than the deliverable. That is not a limitation of this particular service; it is true of machine translation generally, and the honest workflow is machine first, person after.
What Translation can do
What an agent can do with it, and what comes back when it does.
Coverage and limits
What it does not do, said up front, so nobody plans a workflow around something that is out of scope.
Plain text, not documents
It translates text, not files. Layout, styling and document structure are the caller's business, which in practice means the agent pulls out the strings, translates them, and puts them back.
A draft for anything published
Machine translation is good enough to act on and not good enough to sign. Marketing copy, legal wording and anything with your name on it wants a fluent human reader before it ships.
Names and terms drift
Product names, feature names and industry terms get translated when they should have been left alone unless the request says otherwise. Saying which terms to keep is worth the one line it takes.
It translates, it does not summarize
The response is the same text in another language. If you want it shorter or explained, that is the agent's own work, and asking for both at once usually gets you neither cleanly.
What Translation costs
Web services are billed in credits per call, not per token. There is no separate vendor bill to reconcile.
zero translateCharged in credits by the length of the text, and it is the cheapest call in this catalog: a page of prose costs well under a single credit. That is the point of it. Translating everything that arrives is affordable in a way that reserving translation for the important items never was.
Setup and access
Nothing to set up
Nothing to connect. No translation account, no key, no per-character plan. The model is managed by Okou and chosen for cost and speed, so it can change without anything on your side changing.
Who can use it
Translation is its own permission, granted per agent and per person. It is low risk and cheap, and it is still something you hand out rather than something every agent holds by default.
What teams use Translation for
Reading what customers actually wrote
Support messages, reviews and survey answers in a dozen languages, translated on arrival so the team reads all of them instead of the ones they happen to understand.
Working through foreign-language sources
A competitor's site, a local news story, a regulator's notice. Translate first, then decide whether it is worth a closer look, rather than the other way round.
First-pass localization
Strings, subtitles or help articles translated in bulk, then reviewed by a person for the ones that carry the brand. The machine does the volume, the person does the judgement.
When not to use Translation
Skip it as the last step for anything published under your name, and put a fluent reader after it. Skip it for legal or medical wording, where a certified translation is the requirement rather than a preference. And skip it when the agent's own model is already reading the text for another reason, because asking that model to answer in the target language is one step instead of two.
What this is called elsewhere
The same thing goes by several names in the market. If you have shopped for one of these, this is how it maps to what you get here.
Translation API
Text in, translated text out, over an interface a program can call. That is this, with no account or key of yours.
Machine translation
The standard name for what this is: fast, cheap, good enough to act on, and a draft rather than a deliverable for anything published.
Bulk translation
Running translation over everything rather than over a shortlist. The price is what makes that a reasonable default.
Language detection
Working out what language something is in before translating it. It happens automatically here, so nobody has to ask first.
Localization workflow
The pipeline around the translation: extracting strings, translating them, reviewing the ones that matter and putting them back. The agent does the pipeline; this service does one step of it.
Translation compared
Translation vs a translation API of your own
Same job, one less account and one less bill. Character-priced plans also make teams ration translation, which is exactly the behaviour that makes multilingual work fall behind.
Translation vs asking the agent's model to translate
The agent's own model is better at nuance and much more expensive per word. Use it for the paragraph that matters, and this for the thousand messages that need to be legible.
Translation vs a human translator
Not the same product. A person handles brand, register and legal exposure. This handles volume, and handling volume is what leaves a person free for the parts that need them.
The short version
The cheap translation step to put in the middle of a workflow. Machine first for everything, a person after for anything that ships with your name on it.
Frequently asked questions
Do I need a translation account?
No. Okou runs the model and charges the translation in credits. There is no signup and no key.
Which languages are supported?
The common language pairs a business actually encounters. For a rare pair, it is worth checking a sample before a workflow depends on it.
Do I have to say what language the text is in?
No. The source is detected automatically, and you can state it when you already know and want it respected.
What does it cost?
Credits by the length of the text, and it is the cheapest thing in this catalog. A page of prose costs well under one credit.
Is it good enough to publish?
As a draft, yes. As the final text under your brand, have a fluent person read it first. That is true of machine translation everywhere, not just here.
Can I keep product names untranslated?
Yes, by saying so in the request. Left unsaid, names and industry terms sometimes get translated when they should not.
Can it translate a document with formatting?
It translates text. The agent handles the document: pull out the strings, translate them, put them back where they came from.
Which model does it use?
A managed model chosen for speed and cost, and it can change. Nothing on your side depends on which one it is today.
How you ask for Translation
You describe what you want in plain language. The agent works out which service it needs and makes the calls.
“Translate everything that came into support overnight into English and group it by what people are complaining about.”
“Translate our product one-pager into Japanese and Spanish, keeping the product names exactly as they are.”
“This regulator's notice is in German. Translate it, then tell me which parts affect us.”