What Alpi costs — one metric, public bands, no metering

Alpi's pricing in full: a single billing unit (the installed agent), public graduated bands, grandfathered entry prices, founding-customer terms, and implementation services.

May 28, 2026 · pricing · business

Most agent platforms bill for things that require watching you: seats, traces, runs, tokens. Alpi bills for one thing it never has to measure — the installed agent. This post is the full, public pricing picture: the metric, the bands, what counts as installed, and the terms around it. No hidden enterprise rate card, no negotiated discounts.

One metric: the installed agent

The billing unit is a productive profile deployed in production doing real business work. No seats, no traces, no tokens, no execution metering. The customer declares how many agents are installed; we never measure usage.

This isn't a pricing gimmick — the agent is the atomic unit of the platform by architecture. Every profile has its own cryptographic identity, memory, budget, peer list, and capabilities. Charging per installed agent prices the unit the platform itself recognizes, and it requires no telemetry to operate. The same "no surveillance" principle that protects your data also shapes the invoice.

Pricing (2026)

Public, graduated, and the same for everyone. Each band prices the agents within it — marginal pricing, so adding one agent never triggers a tier cliff.

Installed agentsPrice per agent / year
1st agentFree (personal / evaluation, BUSL)
Agents 2–10$10K
Agents 11–30$8K
Agents 31–75$7K
Agents 76–150$5K
151+Custom

The bands fall as you scale on purpose: the more agents you install, the better the per-agent economics, so "build it ourselves on a framework" never wins on cost at scale.

Worked examples (annual license):

DeploymentAgentsAnnual license
Small team5$40K
Single department (e.g. a web factory)15$130K
Multi-department organization50$390K
Large operation150$940K

Early price, not discounts

Alpi doesn't discount. Instead, the published price rises as the platform matures — and every customer keeps the band of the year they entered, for as long as their subscription stays continuous. 2026 is the lowest price Alpi will ever have; entering early is rewarded structurally, not negotiated. Indicative guidance: list prices are reviewed annually and expected to rise 15–25% per year while the platform accumulates capabilities and references.

Founding Customer program

The first five enterprise customers enter as Founding Customers:

In exchange: a public testimonial / case study and reference-call availability. Once filled, the program closes permanently.

What "installed" means

An agent (profile) is installed when it:

Profiles used purely for development, testing, evaluation, or sandboxed experiments are not installed agents. Customers self-declare their count, with the right to an annual audit if requested — count, not content.

Support and services

Standard support is included in every paid band: email support with business-day response, documentation access, and upgrade guidance. Premium SLA is available at 20% of the annual license — defined response times, a named engineer, assisted upgrades (free for Founding Customers for three years).

Software licenses don't transform organizations; implementation does. Professional services are decoupled from the license and scaled to actual need — a team with its own agent-engineering capability may not need them at all.

PackageDurationPriceScope
Bootstrap2–3 weeks$24KSetup, daemon install, one workgroup, team training
Deployment6–8 weeks$60KBootstrap + 3–5 production workgroups, custom skills, stack integration
Transformation3–6 months$120K–300KDeployment + scale to 15–50 agents, deep custom skills, runbook
Custom engineeringper day$2K/daySpecific extensions, integrations, custom ALP work

What else you pay (and to whom)

Beyond license and services, the customer bears LLM API costs (mix providers per profile — premium, mid-tier, or local Ollama — to optimize per workload) and cloud infrastructure wherever the daemon runs. Because Alpi is model-agnostic by architecture, you keep your pricing leverage with model vendors; we have no incentive to push you toward one.

Why this is aligned

For the customer, the license is a small fraction of the human cost it displaces, the bands have no cliffs, success isn't penalized (an agent doing a million turns a year costs the same as one doing a thousand), there's no usage metering and no surprise invoice, and the grandfathered entry price means what you've installed never gets more expensive while you stay.

For us, revenue grows with the customer's adoption — more departments agentized, more installed agents — not with surveillance of their usage. Rising list prices reward early customers instead of margin-destroying discounts, and services revenue declines predictably as the customer internalizes capability. The model is built around being aligned with the transformation, not extracting maximum revenue from a single contract.

Licensing, in one paragraph

The agent core is Business Source License 1.1, and each released version converts to Apache-2.0 four years after its first public release (a rolling window — the four-year-old codebase is always fully open source). Source is available to read, study, and modify from day one; personal, research, and evaluation use are free; commercial production use of current versions needs a license. The conversion is your structural protection: if the publisher disappeared tomorrow, every version ever shipped becomes open source within at most four years. The Alpi Link Protocol is Apache-2.0 from day one.


Prices are 2026 list. Figures here are the published rate card; deployment ROI scenarios elsewhere on this blog are directional, not guaranteed. For governance, compliance, or integration specifics: info@satoshi-ltd.com.

theme