ML Env

Use casesSearch & discovery

Let shoppers search in their own words, not your catalogue's words

Meaning-based search next to your keyword search, so loosely phrased queries find the right products even when no words match your titles.

How it works
Your product titles and descriptions are turned into embeddings once. Each shopper query is embedded live and matched against them, so 'warm jacket for cycling in rain' finds your softshell even though none of those words appear in the title. A reranker then orders the shortlist by actual relevance. It runs as a hybrid: your keyword search still handles exact SKU and brand-name queries, embeddings handle everything phrased loosely.
Data you need
Your product catalogue with reasonable titles and descriptions — a standard product feed export from Shopify, Shopware, Magento or WooCommerce is enough. Thin catalogues with title-only entries limit what semantic matching can do.
What to expect
Good at the queries keyword search misses: descriptive, conversational, or phrased in words your catalogue never uses. Quality is capped by your product data — if descriptions are sparse, the model has nothing to match against. Exact-match queries such as part numbers and model codes still need keyword search, which is why this runs as a hybrid, not a replacement. We make no conversion promises; the honest expectation is fewer obviously wrong result pages.
Where people stay involved
A merchandiser should review sample queries per category before launch and periodically after, and keep the ability to pin or block products for specific queries. Relevance judgements are a business call, not a model call.

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.

Also in search & discovery

All 12 →

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.