One control plane for production agentic AI.
Runtime enforcement, identity, evaluations, HITL, and cryptographic audit evidence in one platform.
Priced for builders. Scaled for enterprises. Designed for governed AI agents in production.
What the cost of ungoverned AI agents can look like
You reconstruct it by hand
When an agent acts up, you piece the story together from logs that were never built to answer the one question that matters: what did it do, and who let it.
It acts without asking
An agent calls a tool you never approved, reaches data it should not, and keeps going while no one is watching. The 2 a.m. agent does not wait for your review.
It fails the audit
The board asks how you know your agents are under control. The regulator asks for proof of human oversight. A screenshot does not survive either one.
For the individual builder. Prototype, evaluate, and learn the platform.
Save $298 vs monthly
Cloud
- 3 active agent identities
- 100K policy decisions / month
- 1 operator seat
- 7-day CAL hot retention
- Community support
For a small product or security team. Real workloads, real evidence.
Save $1,598 vs monthly
Cloud
- 10 active agent identities
- 1M policy decisions / month
- 5 operator seats
- 30-day CAL hot retention
- Basic eval campaigns + regression
- Email support, 48hr response
For production teams running governed agents at scale.
Save $7,998 vs monthly
Cloud, Business Hybrid
- 50 active agent identities
- 5M policy decisions / month
- 15 operator seats
- 90-day CAL hot + 275-day cold retention (1-year total)
- 1 Framework Pack credit / year
- Priority support, 8hr response
Multi-team production deployments at scale.
From $95K / year
Cloud, Hybrid · Self-hosted available as add-on
- Contracted agent identities
- Contracted policy decisions
- Unlimited operator seats
- Up to 5 year CAL retention
- Dedicated CSM
- Air-gapped and regulated deployments from $175K/yr
Save approximately 17% on every paid tier with annual billing.
Regulated / Private Enterprise
From $175K / year
For teams that cannot send telemetry off-premises — defense, finance, healthcare, and public sector. The full platform runs entirely inside your boundary, with the evidence to prove it.
- Self-hosted or air-gapped deployment (included)
- Extended evidence retention (up to 5 years)
- Dedicated implementation engagement
- Regulated industry controls
- Custom contracted SLAs
Building locally? Start with EmberSpark.
EmberSpark is the open-source AI agent framework that shares Ember’s bounded autonomy principles. Apache 2.0 licensed. No account required. No telemetry. Run it on your laptop or your infrastructure.
When you’re ready to govern agents in production, migrate workflows from EmberSpark to Veilfire Cloud for runtime enforcement, evaluations, and cryptographic audit evidence.
Platform Capacity by Tier
Every paid tier includes the full platform. The numbers below describe how much you can run.
Dev Cloud
- Active agent identities
- 3
- Standard policy decisions / month
- 100K
- Operator seats
- 1
- Approver / auditor users
- 1
- CAL hot retention
- 7 days
- Framework Pack credits
- 0
- Support
- Community
Teams
- Active agent identities
- 10
- Standard policy decisions / month
- 1M
- Operator seats
- 5
- Approver / auditor users
- 5
- CAL hot retention
- 30 days
- Framework Pack credits
- 0
- Support
- Email · 48hr response
Business
- Active agent identities
- 50
- Standard policy decisions / month
- 5M
- Operator seats
- 15
- Approver / auditor users
- Unlimited
- CAL hot retention
- 90 days
- Framework Pack credits
- 1 included
- Support
- Priority · 8hr response
Enterprise
- Active agent identities
- Contracted
- Standard policy decisions / month
- Contracted
- Operator seats
- Unlimited
- Approver / auditor users
- Unlimited
- CAL hot retention
- 90 days
- Framework Pack credits
- Negotiated bundle
- Support
- Dedicated CSM
Priced to encourage deep instrumentation.
A policy decision is a runtime check performed during agent operation — tool governance, memory access checks, network egress rules, HITL thresholds, sequence rules, rate limits, and CAL audit emissions.
We price routine decisions cheaply on purpose. Every action boundary you instrument is another layer of evidence and another control surface. Adding more safety checks should never feel like a billable penalty.
| Meter | Overage |
|---|---|
| Standard policy decisions | $250 per additional 1M |
| Advanced detector decisions | $1,500 per additional 1M |
| LLM-based runtime classifier or judge calls | BYOK — your LLM provider bills inference. Veilfire orchestration included in plan quota. |
| Additional operator seat | $75 / month |
| Additional approver or auditor user | $15 / month after plan limit |
| Additional CAL hot retention | Priced per million events per retention period |
| CAL cold archive | Annual add-on, priced per 1M events |
Evaluations are assurance work, not guardrail calls.
Insight evaluations consume model inference, scenario execution, judge passes, telemetry analysis, reporting, and the security expertise embedded in our test libraries. Priced accordingly.
Included by tier
| Tier | Scenarios | Campaigns |
|---|---|---|
| Dev Cloud | 25 / month | 1 basic |
| Teams | 150 / month | 3 |
| Business | 1,000 / month | 10 |
| Enterprise | Contracted | Contracted |
Overage and services
| Standard scenario execution | $8 each |
| Adversarial scenario execution | $35 each |
| Multi-turn agent workflow scenario | $75–$150 each, by depth |
| LLM judge pass | BYOK — your LLM provider bills inference. Veilfire orchestration included in eval scenario quota. |
| Full eval campaign report outside plan | $3,500 per report |
| Custom scenario authoring | $1,500 per scenario, or project-priced |
| Custom adversarial test suite | $35K–$75K fixed fee |
| Managed quarterly eval review | $25K / quarter |
Control & Evidence Packs
Framework-mapped control templates, policy rules, scenario tests, evidence mappings, reports, and operational guidance that support compliance programs.
Packs support compliance work. They do not guarantee compliance.
Foundational Controls
Included with every paid tier
Veilfire Open Safety Spec — bounded autonomy, tool allowlisting, action-boundary checks, memory access control, HITL escalation, agent identity, CAL emission
Framework Evidence Pack
$9,000 / year each
NIST AI RMF, ISO/IEC 42001, SOC 2 + AI, GDPR + AI, OWASP LLM Top 10
EU AI Act Readiness Pack
$15,000 / year
High-risk system readiness, technical documentation support, human oversight mappings, evidence mappings, governance workflows, scenario tests, EU AI Office guidance delta tracking
Canadian AI Governance Readiness Pack
$9,000 / year
PIPEDA, OPC guidance, federal AI directives (Treasury Board Directive on Automated Decision-Making), and proposed AIDA (Bill C-27)
Industry Evidence Pack
$25,000–$40,000 / year each
Healthcare, banking, insurance, public sector, critical infrastructure
Custom Pack Development
$35,000–$100,000 fixed fee
Tenant-specific policies, proprietary controls, internal governance frameworks. Annual maintenance: 25% of build fee.
Bundle discounts
Applied to the annual catalog pack list price subtotal. Excludes Custom Pack Development.
- 3 packs·10% off paid pack subtotal
- 4 packs·15% off paid pack subtotal
- 5 or more packs·20% off paid pack subtotal
Deployment Options
Cloud, hybrid, and self-hosted delivery for multi-team production. Fully air-gapped and regulated deployment is covered by the Regulated & air-gapped tier above.
Standard Enterprise
From $95K / year
- Cloud or Hybrid deployment
- Self-hosted available as +$65K / year add-on
- Premium support
- Multi-team governance scale
Stack the alternative.
Use your own assumptions for the multi-vendor governance stack. Compare to the Veilfire integrated platform. No competitor pricing baked in.
Your multi-vendor stack
Annual subscription for a runtime governance / policy enforcement platform
Third-party guardrail / safety API usage
Evaluation platform, scenario library, judge passes
Vendor setup, SI / consulting engagement
Engineering time to keep multi-vendor stack working
Your stack
$110,000/ year
Veilfire Business
$39,990/ year
You'd save $70,010 / year.
Business + 2 Framework Packs
$61,590/ year
Business + EU AI Act Readiness + SOC 2 + AI (3-pack bundle 10% off).
You'd save $48,410 / year — plus the framework evidence mappings.
This calculator uses your own assumptions. It does not quote competitor pricing. Numbers reflect typical multi-vendor stack components.