How It Works

01 The shape of it

Fig. 01 — The gateway is the seam: left of it nothing changes, right of it everything can.

02 Four layers, all open source, all yours

You'll own every layer.

Each layer is established open-source software. The specific components and their configuration are chosen for your workload, existing infrastructure and operational preferences — and set out in your proposal, under NDA.

01 The gateway One endpoint A single endpoint every team talks to instead of talking to a supplier directly. Routing, failover, spending limits per team, and the point at which switching supplier stops being an engineering programme. Open source

02 The record Every request Every request logged with who asked, what answered, how long it took and what it cost. This is what turns one unexplained invoice into a breakdown you can charge back and defend. Open source

03 The serving layer Your capacity What turns capacity you have bought into work delivered. How well this is configured decides your cost per unit of work, and it is where most self-assembled attempts quietly lose their savings. Open source

04 The models Open-weight + premium Open-weight model families, run on infrastructure in your name, with a capped route to a premium vendor kept open. Which model answers which request is decided against your own work, not a leaderboard — and revisited as releases land. Open source

03 Why open source, specifically

Because the point is that you can fire us.

A managed service only its vendor can operate is not a solution to supplier lock-in. It is a smaller supplier with the same problem.

Everything we deploy is permissively licensed, runs in your accounts, and is maintained by communities that are not us. That is not altruism — it is the only version of this business we think survives contact with a competent CTO, and it is why we charge for operating the arrangement rather than for access to it.

Not in the arrangement

Any Guztia-built software, any proprietary control layer, any margin on your usage, any hosted service of ours that your traffic must pass through.

You hold

The cloud accounts, the infrastructure code, the model weights, the usage records, the documentation.

We hold

A monthly invoice and the pager.

Leaving

Thirty days' notice. Nothing to extract, nothing to rebuild, no data to get back.

04 Where open weights still lose

We are not going to tell you the gap has closed.

It has narrowed a great deal. For the bulk of enterprise work — classification, extraction, summarisation, internal question answering, routine code — a good open-weight model is not distinguishable from a premium API in outcome, while costing a fraction. That is most of your volume, and it is the whole basis of the argument.

It is not all of it. There remain tasks where the premium models are simply better, and pretending otherwise would cost you more than you saved. That is exactly why every arrangement we build keeps a capped route to a premium vendor rather than burning the bridge.

Which of your workloads fall on which side is an empirical question about your work, not a matter of opinion — and answering it is the first thing an assessment does.

Long-horizon agents

Many steps with tool use and self-correction. Premium models hold together noticeably longer.

Novel reasoning

Unfamiliar problems rather than variations on well-covered ground. The gap is real.

Very long context

Advertised window size and quality at the far end of it are two different questions.

Low-resource languages

Coverage varies sharply, and a headline multilingual claim will not tell you where.

Harder multimodal work

Premium APIs remain ahead on difficult document and image reasoning, for now.

THE RECOMMENDATION

Route by difficulty, not by loyalty

Each request goes to the cheapest route that does the job well enough — capacity you control for most of it, a premium vendor under a hard ceiling for what genuinely needs one. Every one of those requests is recorded, so you can see precisely what the escape hatch costs and whether it is shrinking. In the arrangements we run, it shrinks.

What that does to the bill

05 Questions

Why don't you publish the exact components you deploy?

Every component is established, permissively licensed open-source software — that is a commitment, not a secret. Which components, in which configuration, for which of your workloads is a per-client decision, and it is where most of the cost difference is made or lost. It is also an attack-surface disclosure we would rather not make on behalf of every client at once. You receive the full bill of materials and the proposed design in the proposal, under NDA, before any commitment.

Could we build this ourselves?

The components are open source, so in principle yes. What teams underestimate is not the installation — it is choosing which workloads can move without anyone noticing, sizing capacity against demand that has not settled, and the eighteen months afterwards, during which the arrangement has to be kept current and is nobody's job. That judgement and that operation are what we sell.

Is any of this proprietary to Guztia?

No. The components are widely used open-source software that Guztia did not write and does not licence to clients. Everything runs under accounts in the client's own name. If the arrangement ends it keeps working, and another firm or an internal team can take it over.

Where do open-weight models still fall short?

Long-horizon agent tasks, genuinely novel reasoning, quality at the far end of very long context windows, some low-resource languages, and the harder multimodal document tasks. For the bulk of ordinary enterprise work the difference in outcome is usually not detectable. Which of your workloads fall on which side is established in the assessment.