Read a whole tender pack overnight and get a bid / no-bid memo
Reads a whole tender pack overnight and drafts a bid or no-bid memo: what is asked, what would disqualify you, and what is dangerously absent.
- How it works
- You load the entire tender pack — invitation, specification, contract terms, pricing schedules, every annex — and overnight the machine reads all of it in one pass and produces a bid / no-bid memo: what is actually being asked, what would disqualify you, what the real obligations are, and what looks dangerously absent. This is not the lookup tool that answers single questions from a page; here the model reasons across the whole document to reach a recommendation that a person then owns. It runs on ml2, the machine with the bigger card, because a job this size keeps the attention layers and the whole working memory of the read on the GPU while the bulk of the model sits in ordinary system memory — for any one word only a small part of the model does the work, which is what lets it finish a long read overnight. That same design is why it cannot answer while you wait: it is an overnight job by nature, not a chat.
- Data you need
- The complete pack as files with selectable text — the invitation, the terms, the pricing schedules and all the annexes, not a subset. Public-sector packs routinely arrive as scanned or image-only PDFs with no text layer, and those return nothing useful until they have been OCR'd first, as no vision model is served. Nothing of your own is required, but a memo written from half the annexes is worse than none, so a complete, readable pack is the real prerequisite.
- What to expect
- Where the pack is complete and readable, the memo is a strong first read that surfaces the clauses that decide the matter and saves a person from cold-reading hundreds of pages. It will still miss things: a disqualifying requirement buried in an annex the model treated as minor can be left out of the memo entirely, so "nothing dangerously absent" is never proof that nothing is. Contradictions between two documents can come back as one confident recommendation rather than a flagged conflict, and the model cannot price the bid or judge whether you have the capacity to deliver. It reads and drafts; it does not decide, and its reading is not legal advice.
- Where people stay involved
- The bid owner reads the memo against the pack and makes the call — the memo is a briefing, not the decision. Every mandatory requirement and deadline it names is checked against the source document, because in tendering a wrong "yes, we qualify" is a disqualification, not a typo. The go / no-go decision, and any commitment that goes into the submission, is always a person's and never automatic.
Which model, and what it costs to run
Qwen3-235B-A22B-Instruct-2507
No public benchmark ranks this job on our roster — reading a whole document and reasoning across it is exactly what the current long-context tests fail to predict. A job this hard starts from a large model, and we prove it on your own material before anything ships.
- Licence
- Apache-2.0
- Weights at 4-bit
- 132 GB
- Context
- 256K tokens
- Publisher
- Alibaba (Qwen Team)
What the hardware costs
One 141 GB card holds it
- Rent in the EU
- $4.50/hrNebius, Finland (eu-north1)
- Buy the card
- $29,500new, one-off
- Or rent it by the token
- $0.1 / $0.3per 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.
Reading a long document whole
No benchmark measures this
No current benchmark ranks today's open models here. HELMET showed that the popular test — finding a planted sentence — predicts nothing, and its own table has not been rerun on 2026 models.
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.
Sticking to the document
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.
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 sales & b2b
All 15 →Draft a quote from a customer's enquiry email and your price list
Turns an inbound enquiry and your price list into a filled-in quote draft for a salesperson to check and send — a document to correct, not a blank page.
Pull the line items out of RFQ emails and PDFs into a spreadsheet
Extracts part numbers, quantities, units and dates from RFQ emails and PDFs into one consistent table, instead of someone retyping them.
Sort incoming sales emails: real enquiry, existing customer, supplier, junk
Tags every email hitting the sales inbox — new enquiry, existing customer, supplier, junk — within seconds, so real enquiries reach the right person.
Grade new leads against your own definition of a good customer
Scores each new enquiry against your written definition of a good customer, with a short 'why' note the salesperson can read and disagree with.
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.