Inference is electricity.We sell it like a utility.

Your agents run on your own APG appliance. The one thing they need that you may not want to host is the model itself. Solutions Factory will run the model as a utility, metered and itemized. That managed tier, Tier 1, is coming soon. Today you run inference on your own key or your own GPU. There are three tiers, and one config line moves you between them.

Three ways to run it

Pick where the model runs.

Your agents run on your own machine. The model behind them can run through us, on your own key, or on a computer you own.

Through us: Metered and itemized like a utility bill. Not open yet.
Your own key: You hold the key and the provider bills you. No margin from us.
On your machine: The model runs on a computer you own. Nothing leaves your network.

Where the model runs

Your agentsSolutions FactoryModel provider

Where it goes: Your agents send each request to us. We pass it to the model provider.
Who sends the bill: You pay us for what you use, itemized like a utility bill.
Your data: Your records stay in your own database. Only the request itself passes through.
Status: Not open yet. What is kept, and for how long, is stated in writing before it opens.

Where the boundary sits.

Your database, the APG runtime and your own GPU sit inside the infrastructure you control. Inference leaves that boundary only on Tier 1, through the Solutions Factory proxy to the model provider, or on Tier 2, straight to the model provider on your own key. On Tier 3, nothing crosses the boundary at all.

Tier 2 · BYOK, available todayTier 3 · Sovereign, available today

An electric utility buys power at scale and bills you for what you draw, with the rate printed on the bill. That is how Tier 1 will work with inference when it opens. We will buy the capacity, meter what you draw, and show our margin as a line on the bill. If you would rather buy direct or generate your own, the meter comes out and you pay us nothing for inference.

That is the whole model. The rest of this page is the three ways to wire it.

Tier 1 · SF managed

Metered through Solutions Factory, itemized like a power bill.

Tier 1 is not open yet and cannot be bought. When it opens, your APG appliance will send model calls through our servers. We will buy the capacity, meter what you use and show our margin as a line on the invoice. Before anyone can switch it on, we will state in writing what we keep of the text you send and for how long. Your stored data does not pass through us on any tier. Until then, start on Tier 2 or Tier 3. Moving to Tier 1 later is one config edit.

Monthly inference

Illustrative

Model wholesale · 8.0M tokens$24.00

SF infrastructure margin · 30%$7.20

Your retail total$31.20

Illustrative figures only. We confirm the margin percentage before launch, and nothing here is a quoted rate.

The margin is a line you can read on the bill. You will be able to watch your usage, forecast it, and compare it against bringing your own key at any time.

Billing will be monthly and predictable. You will pay for the tokens you draw.
The proxy is designed to retain no prompt content, to be confirmed before launch; your persistent data does not pass through SF on any tier.
There is no commitment to stay. One config edit moves you to BYOK or sovereign.

Tier 2 · BYOK

Bring your own key. We charge for software only.

If you already hold a model contract or buy at enterprise volume, point APG at your own key. Your model provider bills you directly, Solutions Factory takes zero inference margin, and you pay us for the software license and support alone.

apg config

$ apg config set inference.tier byok
$ apg config set inference.api_key $MODEL_API_KEY
inference: byok · billing direct to your model provider · SF margin: none

Tier 3 · Sovereign

Self-host the model on hardware you own.

If you work in a high-privacy field such as legal, healthcare or defense-adjacent work, or you have a GPU budget, run a local model on your own hardware. There is no inference bill from us and none from a model provider. You pay for the hardware and the electricity, and nothing else. Prompt content stays on your network, and the sale is software only.

Specifying and installing that machine is part of a Delivery engagement, which carries a fee of its own. On hardware: run a 7B-or-larger model at Q4 quantization or better for reliable structured output. A single modern GPU with 12 to 16 GB of VRAM handles an 8B model at Q4 comfortably. Larger models and higher quantization want more memory. Below that, structured-output reliability drops, and APG’s validation will flag it.

apg config

$ apg config set inference.tier sovereign
$ apg config set inference.local_model <your-local-model>
inference: sovereign · local model · SF margin: none

Pick by profile, change your mind anytime.

Your situation	Tier	Why
New customer, want the least friction	Tier 2 · BYOK	Available today. Add your own model key and pay your model provider directly. Tier 1, metered through us, is coming soon.
Existing model contract or enterprise volume	Tier 2 · BYOK	Use the rate you already negotiated; pay SF for software only.
Highest-privacy vertical or own GPU budget	Tier 3 · Sovereign	No prompt content leaves your network, and no inference bill from us or from a model provider.

Software and inference, priced separately and plainly.

The pricing page lays out the software-license tiers against the three inference tiers, with no per-seat line anywhere on it.

Talk to us See pricing