We are now part of the NVIDIA Inception Program.Read the announcement
Pricing

Test free, scale sovereign

Run a 30-day pilot on your infrastructure. Upgrade when ready. No hosted cloud, no data egress.

Free POC Pilot

Full Atlas stack on your infrastructure for 30 days. Zero cost, no credit card required.

$0

$0/year

Best for

Proof of ConceptDevelopment TeamsTechnical Evaluation
Deploy Free POC Pilot

Features included:

  • Full Atlas stack (Runtime, Deploy, Serve, Core, Studio)
  • Deploy on your infrastructure
  • OpenAI-compatible API
  • Arabic-first routing (Falcon-7B, JAIS-13B, Qwen-1.5-7B)
  • Infrastructure activity journal & data residency controls
  • Zero external calls by default
  • Email support
  • 30 days, limited program slots
  • Production SLA
  • Priority 24/7 support
MOST POPULAR

Enterprise License

Production deployment on your infrastructure with SLA and 24/7 support. Volume-based pricing as you scale.

From $5k

From $60k/year

Best for

Government AgenciesFinancial ServicesHealthcareDefense
Request Enterprise Quote

Features included:

  • Everything in Free POC Pilot
  • Atlas stack (Runtime, Deploy, Serve, Core, Studio)
  • On-prem or private cloud deployment
  • Zero-trust isolation architecture
  • Unlimited usage on your infrastructure
  • Infrastructure activity journal & data residency controls
  • Air-gapped deployment options
  • 99.5% uptime SLA
  • Priority 24/7 support with account manager
  • Early access to new features (V1.1+)

Comparison

Feature Comparison

See the difference between Free POC Pilot and Sovereign Enterprise

FeatureFree POC PilotSovereign Enterprise
Model AccessFalcon-7B, JAIS-13B, Qwen-1.5-7B, GPT-4Any model, on your GPUs
Data ResidencyManaged region (in-country)100% On-Prem or Private Cloud
DeploymentFree POC Pilot (30 days)Air-Gapped, One-Command Install
Zero-Trust SecurityApplication-levelPatented Runtime (mTLS enforced)
Sovereignty ControlsData residency + no external calls
Activity JournalOptional local log (7 days)Extended local archive
SupportEmail / CommunityDedicated team + SLA 99.5%
Cost ModelFree POC Pilot, then upgradeFixed infrastructure cost

Cost Structure

Total cost of ownership

Compare costs across pilot, production, and scale deployments.

Free POC Pilot

POC (30 days)$0

Full stack on your infrastructure, no credit card

Then upgrade to:Custom

Enterprise license based on scale and support

Infrastructure:You pay provider

GPU/CPU costs billed directly by your cloud

Enterprise License

1 Year (starting at)From $60k

(~$5k–$15k/mo depending on volume + support)

3 YearCustom

(Multi-year discounts available)

Scale CaseContact

(Custom pricing for 5B+ tokens/month)

40-70% cost savings at scale

Organizations processing 1B+ tokens/month can expect 40–70% reduced costs vs public APIs. On-premise deployment means no per-token fees—just infrastructure costs. Plus: No vendor lock-in, data stays on your infrastructure.

Enterprise

Enterprise plans & support

Choose the support level and SLA that matches your sovereignty posture.

Startup

PRICING

$5-7.5k/mo

VOLUME

100-500M/month

Support:Email
Volume:Base pricing
SLA:None

Scale-Up

PRICING

$8-15k/mo

VOLUME

500M-5B/month

Support:Slack + email
Volume:10-15% discount
SLA:99.5% SLA

Enterprise

PRICING

$15k+/mo

VOLUME

5B+/month

Support:Dedicated team
Volume:20-30% discount
SLA:99.9% SLA + 4hr response

FAQ

Frequently Asked Questions

Get clarity on air‑gapped deployments, hardware requirements, and enterprise pricing models.

Need a custom quote?

Talk to an architect and get a deployment‑grade pricing plan.

Contact sales