AIsion / Editions

Priced like a platform.
Delivered like a practice.

Every edition includes all four capabilities and the full governance layer. You choose the scale, the deployment posture and the level of Edge91 delivery support.

Launch
Pilot

Prove value on one high-impact use case in weeks.

  • All 4 capabilities
  • Cloud deployment
  • Up to 25 users · 5 connectors
  • Guided Discovery + Build sprint
  • Standard support
Start a pilot
Sovereign
Custom

Sovereign cloud or air-gapped on-prem for regulated estates.

  • Everything in Scale, plus:
  • Sovereign / on-prem / air-gap
  • Open-weight models, your keys (BYOK)
  • Dedicated delivery pod
  • 24/7 enterprise SLA
Contact us

Pricing is tailored to scope, usage and deployment mode. Every engagement includes Edge91's five-phase delivery — Discovery → Operate.

In every edition

You never buy a feature twice

All 4 capabilities

GenAI, documents, analytics and agents — always included.

Full governance

Audit, HITL, permissions and responsible-AI by default.

Your models

Bring commercial or open-weight LLMs, hosted or private.

Edge91 delivery

A practice, not a download — experts in weeks, not quarters.

Questions

Pricing FAQ

How is AIsion priced?
Pricing combines a platform subscription (scaled to users, connectors and usage) with a delivery engagement sized to your use cases. Because AIsion productizes Edge91's consulting practice, delivery is measured in weeks — you see value before you sign a large multi-year commitment.
Do all editions include every capability?
Yes. GenAI assistants, document intelligence, predictive analytics and agentic automation — plus the full governance layer — are in every edition. Editions differ by scale, deployment mode and support level.
Can we start small and grow?
Absolutely. Most customers begin with a Launch pilot on one high-value use case, prove the ROI, then expand to Scale across teams and domains without re-platforming.
Which LLMs can we use?
AIsion is model-flexible. Use leading commercial models or open-weight models, hosted in your cloud or fully on-prem. You can route different tasks to different models and bring your own keys.
Is our data used to train models?
Never. Your content is grounded through retrieval, not used to train foundation models. In sovereign and on-prem modes, data never leaves your boundary at all.

Find the right edition

Tell us your use case and environment — we'll recommend the fastest path to value.