Simple and transparent pricing

Cloud-hosted SaaS or on-premises — same platform, your terms. Start free with up to 10 seats and scale when you're ready

14-day free trial · No credit card required

Choose Your Deployment

Cloud-hosted SaaS or on-premises — same platform, your terms.

Per seat per month. Start free, scale when ready.

Annual license based on cluster size. Unlimited users and requests within cluster capacity.

Per seat per year – save 20% vs. monthly. Requests pool across your team.

Monthly Annually −20%

Free

$0
Up to 10 seats
  • ✓ AI Gateway with smart routing
    ✓ Cost dashboard with team attribution
    ✓ Token compression
    ✓ Basic guardrails
    ✓ 5 MCP connectors
    ✓ 7-day log retention
    ✓ Community support
Sign Up Free

Starter

$12/seat/mo
Minimum 10 seats
  • Everything in Free, plus:
    ✓ ROI baseline: 3 outcome connectors
    ✓ Advanced filtering & PII detection
    ✓ Basic tracing
    ✓ 10 MCP connectors
    ✓ Role-based access control
    ✓ 30-day log retention
    ✓ Email support · 99.5% SLA
try free

Enterprise

Custom
Flat annual contract · Volume pricing
  • Everything in Business, plus:
    ✓ Team and individual performance insights (governed, audit-logged)
    ✓ Unlimited seats, flat annual contract
    ✓ Custom outcome systems & metrics
    ✓ SSO/SAML (enterprise providers)
    ✓ Unlimited MCP connectors
    ✓ Custom retention & audit export
    ✓ Source code audit
    ✓ Dedicated success manager
    ✓ White-glove onboarding
Talk to Sales

Small Cluster

$6,500/year

Annual license

Up to 5 nodes · 40 vCPUs
  • ✓ AI ROI & Impact — team-level
    ✓ Full AI Gateway + smart routing
    ✓ Full analytics & tracing
    ✓ 100+ MCP connectors
    ✓ Agent registration & control (up to 10 agents)
    ✓ Host up to 3 local models
    ✓ SSO / SAML / LDAP
    ✓ Quarterly updates · Business-hours support
Request Quote

Large Cluster

$85K+/year

Annual license

21+ nodes · Custom vCPUs · Unlimited
  • Everything in Medium, plus:
    ✓ Custom outcome systems & connectors
    ✓ Custom guardrail connection
    ✓ Unlimited registered agents & local models
    ✓ Continuous updates
    ✓ 24/7 dedicated support + engineer
    ✓ White-glove onboarding (4 weeks)
    ✓ Included professional services hours
Talk to Sales

A seat is one human user or one agent type — all agents of the same type share a single seat. Business is priced per seat with no minimum and no cap; Enterprise is a flat annual contract with unlimited seats. Contributors measured by AI ROI & Impact don't consume seats — you only pay for people who sign in. Model costs are passed through at provider rates.

Compare Plans

Full Feature Comparison

See exactly what's included in each plan

Feature Free Starter Business Enterprise
Seats & Usage
Seats Up to 10 10+ Unlimited, flat annual
Measured contributors.
People tracked by ROI who don't sign in
— Unlimited, no seat Unlimited, no seat
AI ROI & Impact
Cost attribution by team & project ✓ ✓ ✓
Outcome connectors (GitHub, Jira, and more) — 3 ✓+ custom systems
Cost-to-outcome join
Spend, quality and delivery signals on one row
— — ✓
Post-hoc quality scoring — — ✓
Custom impact metrics — — ✓
Team-level impact dashboards — — ✓
Individual leaderboards
Off by default · audit-logged · small groups suppressed
— — ✓
Scheduled ROI reports (email, Slack) — — ✓
Impact metric history 7 days 30 days Custom
AI Gateway
Smart routing (LLMs + MCP) ✓ ✓ ✓
Prompt optimization System prompt System prompt per provider Full set
Model output arbitrage — — ✓
Retrieval Augmented Generation — — ✓
AI cost optimization & token compression
Token compression ✓ ✓ ✓
Per-team compression policies — — ✓
Custom compression profiles — — ✓
AI security & Guardrails
Basic guardrails ✓ ✓ ✓
Advanced filtering & PII detection — ✓ ✓
DLP & policy enforcement — — ✓
Custom guardrail connection
Bring your own guardrail engine or classifier
— — ✓
Team & Agent AI Performance
Cost dashboard ✓ ✓ ✓
Basic tracing — ✓ ✓
Full tracing & debugging suite — — ✓
Department-level billing — — ✓
Log retention 7 days 30 days Custom
Audit export — — ✓
AI Agent Control & MCP Connectors
MCP connectors 5 10 Unlimited
Agent registration & control — — ✓
Cost, time & recursion limits per agent — — ✓
Agent anomaly detection (loops, drift) — — ✓
RAG for agents — — ✓
Access & Support
Role-based access control — ✓ ✓
SSO / SAML ✓ ✓ Enterprise providers
Private deployment
Single-tenant environment in your region
— — ✓
Source code audit — — ✓
Uptime SLA — 99.5% 99.99%
Support Community Email Dedicated CSM
Feature Small Cluster Medium Cluster Large Cluster
Infrastructure
Nodes Up to 5 21+
vCPUs 40 Custom
GPU support — ✓
Air-gapped installation — ✓
Users & requests Unlimited Unlimited
AI ROI & Impact
Cost attribution by team & project ✓ ✓
Outcome connectors (GitHub, Jira, and more) ✓ ✓
Cost-to-outcome join ✓ ✓
Post-hoc quality scoring ✓ ✓
Team-level impact dashboards ✓ ✓
Custom impact metrics — ✓
Individual leaderboards
Off by default · audit-logged · small groups suppressed
— ✓
Scheduled ROI reports (email, Slack) ✓ ✓
Impact metric history Unlimited (local) Unlimited (local)
AI Gateway
Smart routing (LLMs + MCP) ✓ ✓
Prompt optimization & arbitrage ✓ ✓
Local models hosted Up to 3 Unlimited
GPU-accelerated inference — ✓
AI cost optimization & token compression
Token compression ✓ ✓
Per-team compression policies ✓ ✓
Custom compression profiles — ✓
AI security & guardrails
Full guardrails & PII detection ✓ ✓
DLP & policy enforcement ✓ ✓
Custom guardrail connection
>Bring your own guardrail engine or classifier
— ✓
SSO / SAML / LDAP ✓ ✓
Team & Agent AI Performance
Full analytics & tracing ✓ ✓
Department-level billing ✓ ✓
Log retention Unlimited (local) Unlimited (local)
Audit export ✓ ✓
AI Agent Control & MCP connectors
MCP connectors All 20+ All + custom
Agent registration & control ✓ ✓
Registered agents 10 Unlimited
Cost, time & recursion limits per agent ✓ ✓
Agent anomaly detection (loops, drift) ✓ ✓
Custom MCP connectors — ✓
Corporate data training ✓ ✓
Deployment & Support
Update frequency Quarterly Continuous
Support Business hours 24/7 dedicated + engineer
Onboarding Self-serve White-glove (4 weeks)
Professional services — Included hours
Eval license 30 days 30 days

See what your AI spend is actually producing

Connect your gateway traffic to the systems where work lands, and put spend next to delivered outcomes on the same row.