Transparent Pricing

AI Training & Plans

All prices in CAD. Services are currently available to Canadian customers only. Every build includes shared sovereign training infrastructure on Canadian-hosted servers. Configure your exact build — base model, tier, scope, and add-ons — in our self-serve configurator.

Build Tiers

Choose a parameter size matched to your deployment scope. Each tier offers three training scopes.

Foundation

1B Parameters

$500 – $1,000

A lightweight model tuned for one well-defined task — a single document type or a single workflow step.

  • Essentials$500
  • Standard$750
  • Comprehensive$1,000

Specialist

7B Parameters

$2,500 – $6,000

A mid-size model for department-level use, handling several related document types and moderate decision logic.

  • Essentials$2,500
  • Standard$4,000
  • Comprehensive$6,000
Most Popular

Professional

32B Parameters

$8,000 – $15,000

A high-capacity model for organization-wide deployment, with multi-step reasoning across large, varied document sets.

  • Essentials$8,000
  • Standard$11,000
  • Comprehensive$15,000

Enterprise

72B Parameters

$35,000 – $75,000

Maximum-capacity model for mission-critical, multi-domain deployments at scale.

  • Essentials$35,000
  • Standard$50,000
  • Comprehensive$75,000

Training Scopes

Essentials

One training pass with a standard evaluation suite.

Standard

Extended training with iterative evaluation and prompt-tuning rounds.

Comprehensive

Multi-round distillation, extended validation, and a written benchmark report.

Add-ons

Optional upgrades applied on top of your base build price.

Dataset Preparation

Bring your own dataset

Included

You provide a cleaned, labeled dataset. We validate formatting and compatibility before training begins.

We source and prepare the data

Tier-based

Our team curates a domain-specific dataset based on the goals you describe.

  • Foundation (1B): +$750
  • Specialist (7B): +$2,500
  • Professional (32B): +$5,000
  • Enterprise (72B): +$12,000

GPU Infrastructure

Shared training cluster

Included

Runs on shared sovereign infrastructure. Included with every build.

Reserved H100 allocation

+$1,500

Dedicated H100 GPU time for faster, scheduled turnaround.

Priority B200 allocation

+$4,000

Priority access to next-generation B200 capacity when available. Equivalent H100 capacity is substituted if unavailable.

Integration Options

Deploy your trained model with the integration level that fits your organization.

No integration

Included

Receive your trained model and deployment files only.

API integration

$4,000

Connect your model to an existing application through an API. Typical range: $3,000 – $8,000.

Custom application

$18,000

A purpose-built interface or workflow app built around your model. Typical range: $10,000 – $30,000.

Enterprise platform

$75,000

A multi-user platform with admin controls, analytics, and custom workflows. Typical range: $50,000 – $150,000.

Supported Base Models

Select from six open-weight foundation models during configuration. Pricing is based on build tier and scope — not the base model selected.

DeepSeek-R1-Distill-32B Llama 3.1 Qwen 2.5 Mistral Small Phi-4 Gemma 2

Ready to configure your build?

Use our self-serve configurator to select your model, tier, scope, and add-ons — then checkout securely via Stripe in Canadian dollars.

Configure Your Model Contact Us

Final pricing is confirmed at checkout based on your configuration. All amounts shown in CAD. Applicable Canadian sales taxes (GST/HST/PST) may be added at checkout. Dataset sourcing, GPU upgrades, and integration fees are additive to the base build price.