Translation built for the way you actually build now.
FluentC was designed for what translation is now: an agent calling an API thousands of times an hour, in the middle of a workflow, expecting an answer back before the user notices.
Running more than one agent? Get the same translation from every one →
curl https://api.fluentc.ai/ai_agent/translate \
-H "Authorization: $FLUENTC_KEY" \
-H "Content-Type: application/json" \
-d '{"input": "Your order has shipped.",
"input_format": "text",
"target_language": "es", "mode": "real-time"}'{
"status": "complete",
"translation": "Tu pedido ha sido enviado.",
"source_language_detected": "en"
}import os, requests
res = requests.post(
"https://api.fluentc.ai/ai_agent/translate",
headers={"Authorization": os.environ["FLUENTC_KEY"]},
json={"input": "Your order has shipped.", "input_format": "text",
"target_language": "es", "mode": "real-time"},
)
print(res.json()["translation"])Tu pedido ha sido enviado.const res = await fetch("https://api.fluentc.ai/ai_agent/translate", {
method: "POST",
headers: { Authorization: process.env.FLUENTC_KEY, "Content-Type": "application/json" },
body: JSON.stringify({
input: "Your order has shipped.", input_format: "text",
target_language: "es", mode: "real-time",
}),
});
console.log((await res.json()).translation);Tu pedido ha sido enviado.{
"mcpServers": {
"fluentc": {
"url": "https://mcp.fluentc.ai",
"headers": { "Authorization": "YOUR_FLUENTC_KEY" }
}
}
}Works where you build
n8n
LangChain
OpenAI
Claude
Cursor- and anything that sends HTTP
Ten agents. Different models. One translation.
When every agent translates for itself, the same sentence comes out ten ways. Put FluentC in the loop and they all return the same one, whichever model is behind them.
How it stays the sameSource textYour order has shipped and will arrive on Thursday.Target: es
Each agent translates for itself
4 versions- Support agent Model ASu pedido ha sido enviado y llegará el jueves.
- Sales agent Model BTu pedido fue enviado y llegará el jueves.
- Onboarding agent Model CHemos enviado su pedido; lo recibirá el jueves.
- Billing agent Model DSu envío está en camino y llegará el jueves.
…and every other agent adds a version of its own.
All agents call FluentC
1 version- Support agent Model ASu pedido ha sido enviado y llegará el jueves.
- Sales agent Model BSu pedido ha sido enviado y llegará el jueves.
- Onboarding agent Model CSu pedido ha sido enviado y llegará el jueves.
- Billing agent Model DSu pedido ha sido enviado y llegará el jueves.
…and every other agent returns the same one.
- One translation
- Any model
- Reviewed once
Same engine. Call it the way your stack already works.
REST API
One endpoint. Send text or HTML and a target language, get a translation back.
api.fluentc.ai/ai_agent/translateMCP server
Give any agent a translate tool. Claude, Cursor, your own.
mcp.fluentc.aiLangChain tool
Official Python tools: translate, detect a language, poll a batch job.
pip install fluentc-langchain-tooln8n node
Drop translation into a workflow between any two steps.
n8n-nodes-fluentcAgent flows for your industry.
Patient message triage
A patient writes in their language. A triage agent works in English. A nurse reviews.
- Patient message
- Triage agent
- FluentC, both ways
- Nurse queue
- Reply sent
With FluentC in the loop
- The triage agent works in English while the patient reads their own language, and every agent gets identical wording for the same instruction.
- A nurse reviews a phrase once. The words that were reviewed are the words every agent sends next.
Fee and dispute questions
A support agent answers customers in their language, using your policy text.
- Customer message
- Support agent
- Policy lookup
- FluentC, both ways
- Reply
With FluentC in the loop
- Fee and policy phrases return identical wording from every agent and every model version.
- Upgrade or replace a model without changing a word customers have already read.
Catalog localization
A product feed is translated overnight and published to every storefront.
- Product feed
- Nightly batch
- FluentC, JSON in and out
- Storefront
- Search index
With FluentC in the loop
- JSON and HTML come back in the same shape, so nothing needs reassembling.
- Unchanged rows return from memory, so a nightly run only translates what changed.
Concierge agent
Guests ask about their stay in their language. Staff and agents work in English.
- Guest question
- Concierge agent
- Itinerary and policies
- FluentC, both ways
- Reply
With FluentC in the loop
- Policy wording is the same in every channel.
- Guests read their own language. Staff read theirs.
UI strings in CI
A pipeline translates locale files on every change and opens a pull request.
- Code change
- CI job
- FluentC, JSON in and out
- Pull request
- Release
With FluentC in the loop
- Keys and structure are untouched. Only string values are translated.
- Unchanged strings come back identical, so diffs show only real changes.
Client intake
Inquiries arrive in any language. An intake agent works in English and replies in kind.
- Inquiry
- Intake agent
- FluentC, both ways
- Attorney queue
- Reply
With FluentC in the loop
- The intake agent works in English while the client reads their own language.
- Detect the language first when it isn't stated.
Work instructions and safety notices
Source procedures are translated in a batch and pushed to every station.
- Source procedures
- Batch job
- FluentC
- Floor tablets
- Sign-off
With FluentC in the loop
- Each safety phrase is identical on every sign, tablet and bot.
- Repeated phrases come from memory, so revisions are quick.
Owner Q&A and service scheduling
An agent answers owners from the manual and books service, in the owner's language.
- Owner question
- Service agent
- Manual and VIN lookup
- FluentC, both ways
- Booking
With FluentC in the loop
- Warning-light and maintenance wording is identical across the app, the bot and the dealer email.
- Owners ask and read in their own language.
Resident services assistant
An assistant answers residents from program rules, in the language they write in.
- Resident question
- Service agent
- Program rules
- FluentC, both ways
- Answer
With FluentC in the loop
- Eligibility and deadline wording is identical in the assistant, on the web and in notices.
- Residents ask and read in their own language.
Field reports to headquarters
Reports in local languages become English summaries for headquarters and donors.
- Field report
- Summary agent
- FluentC
- HQ dashboard
- Donor update
With FluentC in the loop
- Reports in local languages read in English at headquarters.
- The same local term gets the same English each time.
Example flows, not customer stories.
Translation, in the flow of the work.
Ten agents, one translation
When every agent translates for itself, the same sentence comes out ten ways. Here is why that happens, what it costs, and how to make every agent return identical wording.
Batch or real-time? Choosing a mode
Every FluentC request is either real-time or batch. A short guide to which fits which job, how the plans map to modes, and how to poll a batch job without waiting forever.
Giving a support agent a second language
Your support agent works in English. Your customers don't always write in it. How to add detection and translation around the agent, in about thirty lines, without changing the agent.
What is FluentC?
How do I connect it to my agent?
Should I use real-time or batch?
What does “the same translation every time” mean?
How many languages are there?
es, fr and ja. Which of them your key can use depends on your account, and the languages endpoint lists them.What does it cost?
Fluent is the point.
Most translation APIs were designed when "translation" meant a human pasting text into a web form. Output here reads like a native speaker wrote it, and comes back fast enough to sit inside a user-facing request.
- esTu pedido ha sido enviado.
- jaご注文の商品は発送されました。
- deIhre Bestellung wurde versandt.
- frVotre commande a été expédiée.
- ptSeu pedido foi enviado.
- ko주문하신 상품이 발송되었습니다.
- esEspañol
- frFrançais
- deDeutsch
- ptPortuguês
- itItaliano
- nlNederlands
- ja日本語
- ko한국어
- zh中文
- hiहिन्दी
- arالعربية
- heעברית
- ruРусский
- plPolski
- trTürkçe
- viTiếng Việt
- thไทย
- idBahasa Indonesia
- svSvenska
- ukУкраїнська
- elΕλληνικά
- swKiswahili
- bnবাংলা
- taதமிழ்
- and many more
We're the scrappy alternative to the giants.
That's on purpose.