Support customers in languages your team doesn't speak

Translates incoming tickets into your working language and your replies back, keeping product names intact — written support across EU languages.

How it works: Incoming tickets in any EU language are translated into your team's working language, and your replies are translated back, keeping product names and fixed phrases from your glossary intact. Paired with reply drafting, a small Dutch team can handle German, French or Polish customers in writing without hiring native speakers for every language.
Data you need: The message flow itself, plus a short glossary of product names and terms that must never be translated. Most shops do not have one written down yet, but it can be built in an afternoon from the catalogue and past tickets. A folder of past bilingual replies, if any exist, helps calibrate tone.
What to expect: Quality is strongest for major EU languages and noticeably weaker for smaller ones such as Estonian or Maltese — have a native speaker review real samples before switching a language on. Idioms and sarcasm in complaints sometimes translate literally, and without a glossary, product and brand names will occasionally get translated when they should not. For routine messages the results are dependable enough to run on every ticket; emotional or legal ones are a different matter, covered below.
Where people stay involved: For routine messages, spot-checks are enough once you have built trust — asking the model to translate a reply back into your own language is a cheap sanity check. For complaints, legal topics or anything emotional, a human who speaks the language (or a vetted template) should sign off: machine-translated apologies can read as cold or odd.

Which model, and what it costs to run

Qwen3-8B

This job answers while someone waits, so latency comes first: we hold it to a single small card and, within that, take the best independent score on sticking to the document.

Licence: Apache-2.0
Weights at 4-bit: 5 GB
Context: 32K tokens
Publisher: Alibaba (Qwen Team)

What the hardware costs

One 48 GB card holds it

Rent in the EU: $1.60/hrScaleway, Paris (PAR2)
Buy the card: $7,569new, one-off
Or rent it by the token: $0.04 / $0.04per M in / out · DeepInfra

Hardware only, third-party prices from 2026-07. The figure excludes the KV cache, which grows with context length and how many people use it at once — sized properly in a conversation, not guessed here. Renting by the token is cheaper up front; why our customers still self-host is below.

Sticking to the document

Vectara Hallucination Leaderboard · HHEM · 17 of 18 models measured

Measured on public documents, by a model acting as judge. Read it beside the answer rate: the lowest hallucination rates on this board belong to models that simply decline more often.

Shorter is betterFree to serveConditions apply⚠ answered under 95%

Independent measurement · Vectara · board updated May 11, 2026

Structured output and tool calls

Berkeley Function Calling Leaderboard · v4 · 13 of 18 models measured

Measured in a sandbox, on somebody else's functions. It tells you which models are capable of the shape of the job, not which one will survive contact with your API.

Longer is betterFree to serveConditions apply⚠ answered under 95%

Independent measurement · UC Berkeley (Gorilla project) · board updated 2026-04-12

The API is cheaper per token. Here is why our customers don't use it.

We will not pretend otherwise: renting a model by the token from a serverless API costs less per million tokens than a card we run for you. We show that price on every use-case page. What it does not include is the part a shop with a customer database actually pays for.

01
Your data never leaves hardware you can point at
A serverless "we don't retain your data" is a clause in a contract. Running the model on a card in Amsterdam is a fact of architecture: your catalogue, tickets and customer records are never sent to a third party at all. For a GDPR audit, that is the difference between a promise and a floor plan.
02
The price cannot move without your say-so
A serverless rate card is somebody else's lever. The provider can raise the price, retire the model, or change the terms, and your cost moves with it. The same model on the same card costs the same next year — you own the number.
03
The model cannot be taken away
Hosted APIs deprecate models on their own schedule; the one you built on can be gone in a quarter. An open-weight model on your own hardware runs for as long as you keep the lights on. No vendor can end-of-life it out from under you.

And the price gap closes with volume: past a card you keep busy — very roughly four billion tokens a month — owning is cheaper outright, even before the three reasons above.

Getting a case like this one from a conversation to production takes about two months, and you can stop at the end of any phase.

Also in customer service

All 15 →

Support customers in languages your team doesn't speak

Qwen3-8B

Sticking to the document

Structured output and tool calls

The API is cheaper per token. Here is why our customers don't use it.

Also in customer service

Sort the support inbox before anyone reads it

Draft answers to where-is-my-order emails

Suggest replies from your own help pages and policies

Handle return and refund requests against your own policy

Tell us what your team does by hand