ML Env

Use casesCustomer service

Draft answers to where-is-my-order emails

Looks up the real order and tracking status and drafts a reply with the carrier's information, ready for an agent to approve and send.

How it works
When a message is recognised as an order-status question, the system looks up the order in your shop or carrier system. The model then drafts a reply in your tone with the actual tracking status, the delivery window as reported by the carrier, and the next step, and the draft lands in the agent's outbox for a one-click send. Order-status questions are repetitive and answerable from data you already hold, which is what makes them a natural first candidate for drafting.
Data you need
API or export access to order and tracking status: Shopify, WooCommerce, Magento or your OMS, plus carrier tracking via the carrier's API. If tracking today lives only inside a courier's web dashboard, you will need to arrange API access first. Without a machine-readable order status this case does not work — the model must never guess a delivery date.
What to expect
Drafts for the common case — order shipped, tracking exists, one clear question — are usually good and quick to approve. The reply is only as good as your tracking data: if the carrier feed is stale, the model confidently repeats stale information. Split shipments, several open orders for one customer, and returns-in-progress are where drafts will be mediocre and agent review matters most. Emails that bundle a status question with a complaint need a human anyway, and the routing step must catch them.
Where people stay involved
An agent approves each draft before it goes out, at least for the first months. Auto-send is defensible only for the narrowest pattern: order shipped, tracking link exists, and no other question in the email.

Which model, and what it costs to run

Qwen3-8B

This job answers while someone waits, so latency comes first: we hold it to a single small card and, within that, take the best independent score on sticking to the document.

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.

Phi-43.7% · answers 80.7% ⚠Llama 3.3 70B4.1% · answers 99.5%Gemma 3 12B4.4% · answers 97.4%Qwen3-8B4.8% · answers 99.9%Mistral Small 3.25.1% · answers 97.9%Granite 4.0 Small5.2% · answers 100%DeepSeek-V3.25.3% · answers 96.6%Qwen3-14B5.4% · answers 99.9%
Shorter is betterFree to serveConditions apply⚠ answered under 95%

Independent measurement · Vectara · board updated May 11, 2026

Structured output and tool calls

Berkeley Function Calling Leaderboard · v4 · 13 of 18 models measured

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.

GLM-4.672.38%Kimi K259.06%DeepSeek-V3.254.12%Qwen3-32B48.71%Qwen3-235B-A22B47.99%Qwen3-8B42.57%Qwen3-30B-A3B41.39%Qwen3-14B41.03%
Longer is betterFree to serveConditions apply⚠ answered under 95%

Independent measurement · UC Berkeley (Gorilla project) · board updated 2026-04-12

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 customer service

All 15 →

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.