ML Env

Use casesSelling across languages

Let your support team chat with customers in any EU language

Live translation between agent and customer in chat — the customer writes in their language, your agent reads and replies in their own.

How it works
The tool sits between your agent and the live-chat widget. The customer writes in Polish or Portuguese; the agent sees it in English, Dutch or German, replies in their own language, and the customer receives the translation. It runs on the fast small model, which is what makes a live back-and-forth practical rather than a per-message wait. A pinned glossary keeps product and policy terms consistent.
Data you need
The chat stream itself, plus a short glossary of your product names and policy terms in the languages you serve — a one-page list you build once and extend as odd translations show up. Past chat transcripts are useful for checking translation quality on your actual vocabulary before switching it on.
What to expect
Handles major EU languages well; noticeably weaker on smaller languages such as Estonian or Maltese, and on slang and typo-heavy messages. Nuance can flatten — a mildly annoyed customer can read as neutral. The costliest failure is a mistranslated policy statement, so pin refund deadlines and warranty terms in the glossary or use pre-translated templates for them, and tell customers in the chat UI that translations are automated.
Where people stay involved
The agent is the human in the loop by design — they read every incoming message and write every outgoing one. Nothing is auto-sent to the customer. Agents should know the translation is machine-made and ask the customer to rephrase when something reads oddly.

Which model, and what it costs to run

Qwen3-8B

No public benchmark ranks this particular job, so this is not a leaderboard pick — it is where we would start: small, permissively licensed, and strong wherever it has been measured. We prove it on your own content before anything ships.

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.

European languages

openGPT-X European LLM Leaderboard · FLORES-200 translation, BLEU · 10 models

Previous-generation models — the board has not been rerun on the 2026 roster, so read the order of magnitude, not a ranking of today's models. And BLEU scores adequacy, not tone: it cannot tell you whether the register fits your brand. That is judged on your own copy.

EuroLLM-9B25.6 · EULlama 3.1 70B24.3Mixtral 8x7B17.3Mistral-Nemo 12B17.1Teuken-7B sigma17 · EULlama 3.1 8B17Llama 3 8B15.9Teuken-7B v0.615 · EUCommand-R 35B13.8Mistral-7B v0.313.3
Higher BLEU is closer to a human translationEU-sovereign model

Independent measurement · openGPT-X / Fraunhofer IAIS · captured 2026-07-10

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.

  1. 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.

  2. 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.

  3. 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 selling across languages

All 16 →

Next

Tell us what your team does by hand

Describe the process that takes the most time. We will say plainly whether a model is the right tool for it.