Find duplicate products hiding in your catalogue
Embeddings surface near-identical product pairs across the whole catalogue; the model judges each pair and hands you a ranked merge queue to confirm.
- How it works
- Years of supplier feeds leave the same product listed twice under different titles. BGE-M3 embeddings plus the reranker find near-identical pairs across the whole catalogue; the small LLM then reads each candidate pair and judges 'same product', 'variant of same product' or 'different', with a one-line reason. You get a ranked merge queue instead of a manual eyeballing exercise.
- Data you need
- A full catalogue export with titles, descriptions, attributes, and identifiers where present (EAN/GTIN). If you have reliable EANs, exact-match dedupe should run first in plain code — the model is for the records where identifiers are missing or inconsistent.
- What to expect
- Turns an open-ended clean-up into a worked queue, and the one-line reasons make each decision quick to confirm. The known weak spot: products that differ only in a number — 100 ml versus 200 ml, 4-pack versus 6-pack — look almost identical to embeddings. The pair-judgement step catches most of these, but pack-size traps are exactly where you should review hardest.
- Where people stay involved
- A person confirms every merge. Merging two records that are actually different products, or different bundle sizes, damages orders and stock — this is decision support, not auto-merge.
Which model, and what it costs to run
Qwen3-32B
This job runs in bulk rather than to a waiting person, so size is not the constraint — we take the strongest independent score on structured output and tool calls that we may serve freely and that fits on a single card.
- Licence
- Apache-2.0
- Weights at 4-bit
- 18 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
- $0.12 / $0.12per M in / out · Nebius AI Studio · EU
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.
Structured output and tool calls
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.
Independent measurement · UC Berkeley (Gorilla project) · board updated 2026-04-12
Writing in your voice
No benchmark measures this
There is no benchmark for this and there is unlikely ever to be one. Anyone who shows you a chart ranking models on marketing copy has drawn it from a model's opinion of another model.
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.
- 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 catalogue content
All 14 →Turn supplier spec sheets into first-draft product descriptions
Drafts product descriptions overnight from the supplier data you already hold, in your shop's voice — a review job instead of a from-scratch writing job.
Pull sizes, materials, colours and other attributes out of messy supplier text
Reads free-text supplier info and fills a fixed field list — material, dimensions, colour, EAN — returning structured rows for your PIM or shop admin.
Sort new products into the right category in your shop
Embeddings shortlist the most plausible categories for each new product; a small model picks one and explains why — workable even on large category trees.
Write page titles and meta descriptions for every product and category page
Drafts an SEO title and meta description for every page from real product attributes, within your character limits — economical even at long-tail scale.
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.