Where the cost actually lives.

GaaS separates governance intelligence from inference compute. Four of five pipeline stages are deterministic. One requires LLM inference. The architecture separates them so you only pay for what you need.

Start Free Shadow Mode

Free tier $0 — 1K actions/mo · No credit card

✓$0 free tier — 1K actions/mo ✓No credit card, no contract See plans ↓
Architecture

Two planes. One governance brain.

The Control Plane runs on GaaS infrastructure — always. The Inference Plane powers deliberation and is included in every plan. This separation is why the governance tax is proportional to the governance need.

Runs on GaaS.is
Control Plane
Governance logic, regulatory intelligence, policy evaluation, audit infrastructure, and learning systems. Deterministic. This is what your plan pays for.
Routine action latency
< 100ms — deterministic compute
Included in every plan
Inference Plane
LLM compute that powers deliberation agents. GaaS sends structured prompts; the panel produces assessments. Only triggered for high-risk actions — 5–15% of total volume.
Deliberation latency
200ms–5s — multi-agent panel
Pipeline Anatomy

Where the time goes, stage by stage.

Most actions never leave the Control Plane. When they do, the deliberation stage is isolated and bounded.

Routine path
85–95% of actions · No deliberation triggered
Intent Declaration
schema parse
< 5ms
Enrichment
parallel API fetches
< 50ms
Policy Evaluation
rule matching
< 10ms
Decision + Audit
hash chain + store
< 25ms
Total / action
All Control Plane · end to end
< 100ms
Deliberation path
5–15% of actions · Multi-agent panel triggered
Round 1 (parallel)
4–6 agent assessments, concurrent
Round 2 (cross-exam)
dissenters respond to each other
Round 3 (final vote)
revised positions + final votes
Total / deliberation
Inference included in every plan
200ms–5s
Bills as one decision. A completed deliberation is one deliberation-class decision ($0.05) beyond your included actions. A failed deliberation bills as routine — you don’t pay for our failures.
What you actually pay per governed action
Blended across all pipeline paths · Full governance included
Starter
50K actions/mo
$0.01
per governed action
Growth
500K actions/mo
$0.005
per governed action
Enterprise
5M+ actions/mo
Custom
volume-optimized
Full regulatory assessment, policy evaluation, immutable audit trail, and multi-agent deliberation when triggered — for less than a cent per action at scale. The platform fee pays for governance intelligence, not compute markup.
Pricing

Two ways to run it. One governance brain.

Every plan runs the full pipeline — control plane, CKB, audit trail, learning engine, and dashboard. Choose Complete and deliberation inference is included; choose Core and deliberation runs on your own model endpoint for 30% less platform fee.

Plans
Full pipeline, fully managed
We handle everything — including the LLM inference for deliberation. Powered by Claude Opus 4.6, the latest frontier model with a 1 million token context window. No model configuration needed.
Free
1K actions · 1 agent · Verified email
$0/mo
Developer
10K actions · 3 agents · Monthly
$99/mo
Starter
50K actions · 3 agents · Monthly
$500/mo
Growth
500K actions · 15 agents · Monthly
$2,500/mo
Enterprise
5M+ actions · Unlimited · Annual
Custom
Everything included
  • ✓ Control Plane
  • ✓ Deliberation inference
  • ✓ CKB access
  • ✓ Learning Engine
  • ✓ Audit trail (90-day hot)
  • ✓ Compliance Dashboard
  • ✓ Shadow mode (14 days)
  • ✓ Standard onboarding
GaaS Core · BYOM
Same brain, your inference
Point deliberation at your own model endpoint — your provider, your rates, your data residency. You run the same control plane, CKB, audit trail, and learning engine; we bill orchestration only, so the platform fee drops 30%.
Starter
50K actions · 3 agents · Monthly
$350/mo
Growth
500K actions · 15 agents · Monthly
$1,750/mo
Enterprise
5M+ actions · Unlimited · Annual
Custom
Everything included, except inference
  • ✓ Control Plane
  • ✓ Your model endpoint
  • ✓ CKB access
  • ✓ Learning Engine
  • ✓ Audit trail (90-day hot)
  • ✓ Compliance Dashboard
  • ✓ Shadow mode (14 days)
  • ✓ Standard onboarding

Core starts at Starter — the Free and Developer plans are managed-inference only. Deliberation tokens are billed by your own provider.

Plan Comparison

Everything, side by side

All plans include the full governance stack. Volume and agent seats scale with your usage. Core mirrors Complete from Starter upward — the Free and Developer plans are managed-inference only.

All plans
Full pipeline — deliberation inference included
Free $0 Developer $99/mo Most Popular Starter $500/mo Growth $2,500/mo Enterprise Custom
Volume
Included actions/mo 1K 10K 50K 500K 5M+
Agent seats 1 3 3 15 Unlimited
Governance Stack
Control Plane ✓ ✓ ✓ ✓ ✓
Managed deliberation ✓ ✓ ✓ ✓ ✓
CKB access ✓ ✓ ✓ ✓ ✓
Learning Engine ✓ ✓ ✓ ✓ ✓
Audit trail ✓ ✓ ✓ ✓ ✓
Compliance Dashboard — ✓ ✓ ✓ ✓
Operations
Shadow mode 14 days 14 days 14 days 14 days 14 days
Support Docs Email Email Priority Dedicated
Billing Free Monthly Monthly Monthly Annual
Get started Free $99/mo $500/mo $2,500/mo Contact us
GaaS Core — BYOM
Bring your own model for deliberation
Starter $350/mo Most Popular Growth $1,750/mo Enterprise Custom
Volume
Included actions/mo 50K 500K 5M+
Agent seats 3 15 Unlimited
Governance Stack
Control Plane ✓ ✓ ✓
BYOM deliberation ✓ ✓ ✓
CKB access ✓ ✓ ✓
Learning Engine ✓ ✓ ✓
Audit trail ✓ ✓ ✓
Compliance Dashboard ✓ ✓ ✓
Operations
Shadow mode 14 days 14 days 14 days
Support Email Priority Dedicated
Billing Monthly Monthly Annual
Get started $350/mo $1,750/mo Contact us
Overage pricing
Beyond your plan’s included monthly actions. Each decision bills once, in its highest applicable class.
Pipeline Path GaaS Complete GaaS Core (BYOM) % of Actions
ROUTINE Approve / block / modify $0.002 $0.0015 85–95%
DELIBERATION Multi-agent panel $0.05 $0.01 + your tokens 5–15%
ESCALATION Human review routing $0.25 $0.25 1–3%
Fairness rules, stated plainly. Shadow and test decisions are never metered. A failed deliberation bills as routine. Each decision counts exactly once, in its highest applicable class. The audit record is not metered and never will be.

Overage billing is active as of 1 August 2026, following a full report-only cycle. Decisions beyond your plan’s included monthly actions bill at the prices above. Your included allowance is netted out first and is never metered, and usage recorded during the report-only period was not charged retroactively.
↘
LLM costs are falling, governance costs aren’t. Claude Opus 4.6 — the latest frontier model with a 1M context window — launched February 2026. GPT-4o is 83% cheaper than GPT-4. The Control Plane (where the governance intelligence lives) is deterministic compute — it doesn’t change with model pricing. As inference costs drop, the deliberation range only gets more favorable. Your costs go down; the governance value stays the same.
The published rates are the rates. On Complete, overage prices are flat per decision — no token math, no model-pricing pass-through, no surprises. On Core, our rates are flat too; the only variable is what your own provider charges for deliberation tokens, billed by them directly. Full billing mechanics are documented at gaas.to.
Pricing Calculator

Model your governance costs.

Configure your expected volume, agent count, and risk profile. See exactly what governance costs — stage by stage, path by path.

Configuration
Pricing model
GaaS Complete
Fully managed
We handle deliberation inference
GaaS Core · BYOM
Bring your model
Same brain, your inference
Monthly governed actions 75,000
1K 25K 75K 250K 750K 2M 10M
Governed agents 3
1 15 30 50
Deliberation rate ?% of actions requiring multi-agent deliberation. Typical: 5–15%. 10%
0% 10% 20% 30%
Escalation rate ?% of actions requiring human review. Typical: 1–3%. 2%
0% 3% 5% 10%
Action path distribution
Routine 88%
Deliberation 10%
Escalation 2%
GaaS Complete
Growth
Estimated monthly
$2,500
/month
Cost breakdown
Platform fee
Growth tier · 500K included actions
$2,500
vs. self-governance (token costs only)
Self-governed token cost ~$12,500/mo
Context tokens recovered ~7.5B tokens/mo
Effective cost per action $0.010
Overage rates per decision
Beyond your plan’s included actions
Routine approve / block / modify $0.002
Deliberation multi-agent panel $0.05
Escalation human review routing $0.25
Each decision counts once highest class
Never metered
Shadow & test calibrate freely $0
Failed deliberation bills as routine $0.002
The audit record free forever $0
Metering status active
What You Get

The platform fee pays for intelligence — not compute.

GaaS isn't an inference proxy. It's cognitive offloading — a governance brain built on proprietary regulatory intelligence, calibrated deliberation, and immutable audit infrastructure.

⍑
Compliance Knowledge Base
12 regulatory frameworks, 60+ policies. HIPAA, GDPR, EU AI Act, NIST 800-53, FedRAMP, CMMC, NIST CSF, SOC 2, PCI DSS, and more. Critical updates propagate in minutes.
⚖
Calibrated Deliberation
Six specialized agent roles with curated prompts, weighted consensus, and cross-examination. Learning Engine calibrates weights daily.
⚙
4-Tier Policy Engine
Universal ethical constraints → regulatory compliance → organizational rules → situational policies. Evaluated in under 10ms.
⚣
Immutable Audit Trail
Hash-chained with DLT anchoring and governance proof tokens. 11-nines durable. Not just what happened — why it happened, who decided, what the alternatives were. Co-signed, non-repudiable, policy-epoch-versioned.
⏳
Learning Engine
Continuous refinement from every governance decision. Override patterns, false positive detection, escalation outcome analysis.
☘
Context Enrichment
Discovers what the agent doesn't know — regulatory context, entity state, prior history. Parallel fetch in under 50ms.
Built for mission, not margin? Nonprofits, NGOs, and veteran-owned businesses qualify for free governance for life. Contact us →
// FAQ

Frequently Asked Questions

Both run the identical governance pipeline — control plane, compliance knowledge base, audit trail, learning engine, and dashboard. Complete includes deliberation inference: we run the LLM compute for multi-agent panels. Core (BYOM) points deliberation at your own model endpoint, so you keep your provider, your rates, and your data residency — and the platform fee drops 30% ($350/mo Starter and $1,750/mo Growth, versus $500 and $2,500). Core starts at Starter; the Free and Developer plans are managed-inference only. On Core you also pay your own provider for deliberation tokens.

Pricing moves to per-decision usage billing scaled by action complexity and risk tier. Low-risk routine decisions are a fraction of a cent; high-stakes decisions with full deliberation are priced proportionately. You will never be charged without notice.

Yes, every governed action consumes one decision. You control which actions route through GaaS — only the consequential ones (payments, communications, record access). Most deployments govern a targeted subset, so your decision budget goes significantly further than it sounds.

No. Shadow Mode starts with just an email and a description of your agents. No credit card, no commitment, no contract.

Both use the same per-decision meter. Shadow Mode evaluates actions without enforcing decisions — it shows what GaaS would have done without affecting your operations. Activating enforcement uses the same cost structure, but decisions are actually applied.

Yes. Nonprofits, NGOs, and veteran-owned businesses qualify for free governance for life — contact sales@gaas.is with your organization details. Early-stage startups building on GaaS are encouraged to reach out about extended access terms.

Start with Shadow Mode

Full pipeline, real actions, zero enforcement. See what your agents are doing — then decide when to go live.

Start Free Shadow Mode →
Free tier $0 — 1K actions/mo · No credit card · Terms
Coming soon