Summarise why customers cancelled, churned or said no

Reads cancellation comments, lost-deal emails and offboarding notes together and writes a periodic brief on the stated reasons.

How it works: Cancellation-form comments, 'sorry, we went with someone else' emails and offboarding-call notes are read together by the large model, which groups the stated reasons — price, missing feature, service incident, no longer needed — and writes a periodic brief with illustrative quotes. For a subscription business or a B2B shop, this is the closest thing to a churn analysis that free text can support.
Data you need: Wherever cancellation reasons actually get written down: cancellation-flow text fields, lost-deal notes in a CRM, offboarding emails. Honest weakness: in many small businesses this data is thin or nonexistent — if churned customers just silently stop ordering, there is nothing to summarise, and adding a one-question exit survey comes first.
What to expect: Stated reasons are polite fictions often enough — 'budget cuts' meaning 'your onboarding annoyed us' — and they systematically understate service failures and overstate price, so the brief is a starting point for conversations, not a conclusion. Volumes are small at this scale: one quarter's 'trend' can be a handful of customers, and the model will not flag that. Personal data in quotes must be removed before the brief circulates — GDPR applies to internal documents too, and the model's own redaction will miss things, so a human spot check is part of the workflow.
Where people stay involved: Whoever owns retention or sales reads the brief against what they know about each account, and the same reader checks quotes for identifying details before the brief circulates.

Which model, and what it costs to run

Qwen3-8B

This job runs in bulk rather than to a waiting person, so size is not the constraint — we take the strongest independent score on sticking to the document that we may serve freely and that fits on a single card.

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

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 voice of customer

All 15 →

Summarise why customers cancelled, churned or said no

Qwen3-8B

Sticking to the document

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

Also in voice of customer

Turn a quarter's worth of product reviews into a short themes report

Tag every review and ticket with your own fixed label list, so you can finally count things

Summarise what your NPS promoters and detractors actually wrote

A Monday-morning digest of what customers complained about last week

Tell us what your team does by hand