ML Env

Use casesSearch & discovery

Turn your help centre into a question-answering assistant with cited sources

Customers ask in their own words; the assistant answers from your help-centre articles and links the article each answer came from.

How it works
The customer asks in their own words, retrieval finds the relevant passages in your existing help-centre articles, and the large language model writes a concise answer with a link to the article it came from. The same index can pre-fill suggested answers when a support email arrives, so your agent starts from a draft rather than a blank page.
Data you need
A real help-centre or FAQ corpus — at minimum a few dozen substantive articles. If you have ten FAQ entries, a plain FAQ page is the better product; don't build this.
What to expect
Expect it to deflect the repetitive general questions — shipping, returns policy, how-to — not the emotional or account-specific ones, which still land with your team. Answer quality mirrors article quality: contradictory or outdated articles produce contradictory answers, so the corpus needs an owner. It answers general questions well and account-specific questions badly unless order data is also wired in, which is a separate project.
Where people stay involved
A clear handoff to a human for anything about a specific order, refunds or complaints. Someone reviews the answer log in the early weeks to catch articles that need rewriting.

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.

Search over your own catalogue

No benchmark measures this

MTEB and BEIR rank embedding models, but their evaluation sets sit inside the models' training data. Scores fall by double-digit nDCG on a private corpus, so the ranking does not transfer to your catalogue.

So we do not show a chart here. We measure it on your own content, in the first week, and you see the result before anything ships.

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.

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.