Community vs Evaluation vs Enterprise

AxonFlow is available in three tiers: Community for self-hosted experimentation, Evaluation (free) for serious evaluation and pre-production validation, and Enterprise for organizations requiring IT, security, and compliance approval.

Who this page is for

This matrix is intended for:

• Security and compliance reviews
• Platform and infrastructure architects
• Procurement and vendor evaluation

If you're evaluating AxonFlow for local development or experimentation, the Community tier covers everything needed to build and validate workflows. When you're ready for a deeper evaluation with more capacity, the Evaluation tier is free.

How to read this table

Community is intended for single-team development and evaluation — no license required. Evaluation provides elevated resource limits for serious evaluation — free license via getaxonflow.com/evaluation-license. Requires a named owner. Enterprise capabilities address organizational requirements such as identity, compliance, multi-tenancy, and procurement.

---

Quick Decision Guide

You want to...	Tier
Prototype AI governance locally	Community
Validate policies before production	Community
Test PII detection patterns (Aadhaar, PAN, SSN)	Community
Test Singapore PII detection (NRIC, FIN, UEN)	Community
Evaluate with more policies and connectors	Evaluation (free)
Use organization-wide policies across tenants	Evaluation (free)
Keep 14-day audit retention for evaluation reviews	Evaluation (free)
Govern images with default safety policies	Community (with MEDIA_GOVERNANCE_ENABLED=true) or Evaluation
Simulate policy changes before deploying	Evaluation (free)
Test policy impact across sample inputs	Evaluation (free)
Route high-risk decisions to human approval	Evaluation (free)
Export evidence packs for internal reviews	Evaluation (free)
Roll out to real users with IT approval	Enterprise
Meet regulatory requirements (EU AI Act, SEBI, MAS FEAT, EBA ML)	Enterprise
Prove compliance to regulators with audit exports	Enterprise
Centralize identity management (SSO, SCIM)	Enterprise

Why Teams Choose Evaluation

Evaluation is for teams that need more capacity during time-bound evaluation and pre-production validation. It's free — register with a named owner to receive your key.

What Evaluation Adds Over Community

Capability	Community	Evaluation
Tenant policies	20	50
Organization-tier policies	Not available	5
Connectors with custom policies	2	5
LLM providers	2	3
Audit log retention	3 days	14 days
MAP plans	25	100
Versions per plan	10	25
Execution history	50	500
Concurrent executions	5	25
HITL Approval Gates	Not available	Max 100 pending, 24h expiry
Policy Simulation	Not available	300/day
Impact Report	Not available	50 inputs/run
Evidence Export	Not available	14-day window, 5K records, 3/day

Everything else in Evaluation works identically to Community — same policy engine, same API surface, same SDK compatibility.

Why Teams Upgrade to Enterprise

Enterprise isn't about more features — it's about what IT, security, and compliance require before approving production deployment.

Identity & Organization Controls

Requirement	Why It Matters
SSO + SAML	IT won't approve systems that bypass corporate identity
SCIM user provisioning	Automated onboarding/offboarding through your IdP
SCIM group-to-role sync	Permissions follow your org structure automatically
Multi-tenant isolation	Required when multiple teams share infrastructure

Compliance & Risk

Requirement	Why It Matters
EU AI Act conformity workflows	High-risk AI systems require documented conformity assessments
10-year audit retention	EU AI Act mandates long-term traceability
SEBI/RBI compliance exports	Indian financial regulators require specific formats
5-year audit retention	SEBI mandate for financial AI systems
MAS FEAT AI System Registry	Singapore FIs must register AI systems with risk ratings
MAS FEAT assessment workflows	Fairness, Ethics, Accountability, Transparency assessments
EBA ML Model Registry	EU banks must track ML models with IRB classification
EBA Model validation workflows	CRR requires documented validation for credit risk models
Human-in-the-Loop (HITL) queue	Regulators require human review for high-stakes decisions
Emergency circuit breaker	RBI FREE-AI Framework requires immediate shutdown capability

Platform & Operations

Requirement	Why It Matters
One-click CloudFormation	Production-grade AWS deployment without DevOps overhead
Usage analytics	Cost attribution across teams and projects
Customer Portal UI	Runtime management without code changes
Priority support + SLA	Required for business-critical systems

---

Summary Comparison

A high-level overview by category before diving into the full matrix.

Category	Community	Evaluation	Enterprise
Policy Engine	Shared policy engine, 63 system policies (view-only), tenant-tier policies (20 limit), sub-10ms enforcement	All Community features + 50 tenant policies, 5 organization-tier policies, policy simulation (300/day), impact reports	All Evaluation features + system policy overrides, unlimited tenant policies, unlimited org policies, full version history, HITL approval queue
LLM Providers	OpenAI, Azure OpenAI, Anthropic, Google Gemini, Ollama with multi-provider failover and weighted routing	Same as Community (up to 3 providers)	All Community providers + AWS Bedrock, cost-optimized routing, runtime weight updates, usage forecasting, auto-downgrade to cheaper models
Connectors	PostgreSQL, MySQL, MongoDB, Redis, HTTP/REST, Cassandra, S3, Azure Blob, GCS (2 with custom policies)	Same connectors (5 with custom policies)	All Community connectors + Salesforce, Snowflake, Slack, Amadeus, HubSpot, Jira, ServiceNow, ML anomaly detection (unlimited custom policies)
Authentication	API key authentication (client ID/secret)	Same as Community	SSO/SAML, SCIM 2.0 user provisioning, SCIM group-to-role sync, multi-tenant isolation
Compliance	Basic PII detection (SSN, credit card, PAN, Aadhaar, Singapore NRIC/FIN), SQLi scanning, audit logging (3-day retention)	Same detection, 14-day audit retention, HITL approval gates (100 pending), evidence export (watermarked)	EU AI Act conformity workflows, SEBI/RBI compliance exports, MAS FEAT assessment workflows, EBA ML model registry, 5-10 year audit retention, HITL queue, emergency circuit breaker, unlimited evidence export
Monitoring	Usage tracking (tokens, cost), budget limits and alerts, decision chain tracing, transparency headers	Same as Community	Usage forecasting, billing export, cost dashboard, execution analytics, compliance dashboards, alert channels (Slack, email, webhook)
Support	Community support (GitHub Issues)	Community support (GitHub Issues)	Priority support with SLA, dedicated success manager, Customer Portal UI, architecture review sessions

Key Enterprise-Only Features

• SSO/SAML + SCIM: Integrate with your corporate identity provider (Okta, Azure AD, etc.). SCIM automates user provisioning and group-to-role mapping so permissions follow your org structure.
• Human-in-the-Loop (HITL) Queue: Route high-risk AI decisions to human reviewers via the Customer Portal. Required by EU AI Act for high-risk AI systems and RBI for certain financial decisions.
• Emergency Circuit Breaker: Immediately halt all AI processing across the platform with a single API call or Portal button. Required by the RBI FREE-AI Framework.
• AWS Bedrock Provider: Native integration with AWS Bedrock for enterprise LLM access within your VPC, enabling data sovereignty and compliance with regional data residency requirements.
• Cost-Optimized Routing: Automatically route queries to the cheapest LLM provider that meets quality thresholds, with auto-downgrade when budgets are approaching limits.
• Compliance Export Formats: Generate audit reports in regulator-specific formats (EU AI Act, SEBI, MAS FEAT, EBA ML) with configurable retention periods up to 10 years.
• Customer Portal UI: Web-based management console for policies, connectors, LLM providers, executions, and compliance -- no code changes required for runtime configuration.

---

Full Feature Matrix

This detailed breakdown is organized into 6 logical groups for easier navigation.

1. Governance & Security

Core policy enforcement, threat detection, and code governance capabilities.

Feature	Community	Evaluation	Enterprise
Policy Enforcement
Policy enforcement engine	✅	✅	✅
Sub-10ms inline governance1	✅	✅	✅
Static Policy API (list, get, CRUD)	✅	✅	✅
Dynamic policy CRUD API	✅	✅	✅
Pattern testing API	✅	✅	✅
Threat Detection
PII detection (SSN, credit cards, PAN, Aadhaar)	✅	✅	✅
SQLi response scanning (basic)	✅	✅	✅
Advanced SQLi detection (ML-assisted)	❌	❌	✅
Code Governance
Code artifact detection in LLM responses	✅	✅	✅
Language detection (14 languages)	✅	✅	✅
Code type categorization	✅	✅	✅
Secret pattern detection	✅	✅	✅
Unsafe code pattern detection	✅	✅	✅
Git provider integration (GitHub, GitLab, Bitbucket)	❌	❌	✅
PR creation from LLM-generated code	❌	❌	✅
Code governance dashboard	❌	❌	✅
Media Governance
Image validation (format, size, dimensions)	✅	✅	✅
OCR-based PII detection	✅	✅	✅
SHA-256 audit hashing	✅	✅	✅
System media policies (5 default rules)	When enabled	✅	✅
Toggle system media policies	✅	✅	✅
Modify system media policy actions/priority	❌	❌	✅
Per-tenant media governance config	❌	❌	✅
Configurable enforcement actions (custom policies)	❌	❌	✅
Cloud analyzers (Rekognition, Vision, Azure)	❌	❌	✅
Face/biometric detection	❌	❌	✅
NSFW/content safety scoring	❌	❌	✅
Document classification	❌	❌	✅
Custom analyzer plugins	❌	❌	✅
Full media audit trail	❌	❌	✅
Identity & Access
SSO/SAML integration	❌	❌	✅
SCIM 2.0 user provisioning	❌	❌	✅
SCIM group sync & role mapping	❌	❌	✅
Evaluation & Simulation
HITL Approval Gates	—	Max 100 pending, 24h expiry	Unlimited, configurable
Policy Simulation (dry-run)	—	300/day	Unlimited
Impact Report	—	50 inputs/run	100 inputs/run
Resource Limits
Tenant policies	20	50	Unlimited
Organization policies	0	5	Unlimited
Connectors with custom policies	2	5	Unlimited
LLM providers	2	3	Unlimited
Audit log retention	3 days	14 days	10 years
MAP plans	25	100	Unlimited
Versions per plan	10	25	Unlimited
Execution history	50	500	Unlimited
Concurrent executions	5	25	Unlimited

2. Audit & Compliance

Audit logging, regulatory compliance, and retention capabilities.

Feature	Community	Evaluation	Enterprise
Audit Logging
Audit logging (LLM calls)	✅	✅	✅
MCP query audit logging	✅	✅	✅
Audit logging for code artifacts	✅	✅	✅
Decision chain tracing	✅	✅	✅
Transparency headers (X-AI-*)	✅	✅	✅
Audit retention	3 days	14 days	3650 days (configurable)
Evidence Export
Evidence Export Pack	—	14-day window, 5K records, 3/day, watermark	Unlimited, clean
Evidence Summary	—	✅	✅
EU AI Act Compliance
Conformity assessment workflow	❌	❌	✅
Human-in-the-Loop (HITL) queue	❌	❌	✅
Emergency circuit breaker	❌	❌	✅
Accuracy metrics & bias detection	❌	❌	✅
10-year audit retention	❌	❌	✅
EU AI Act export format	❌	❌	✅
India Compliance (SEBI/RBI)
India PII detection (Aadhaar, PAN, UPI)	✅ Pattern-based	✅ Pattern-based	✅ With checksum
SEBI AI/ML Guidelines - basic detection	✅	✅	✅
SEBI compliance module (export, 5-year retention)	❌	❌	✅
RBI FREE-AI Framework (kill switch, board reports)	❌	❌	✅
Compliance dashboard	❌	❌	✅
Singapore Compliance (MAS FEAT)
Singapore PII detection (NRIC, FIN, UEN, Phone, Postal)	✅ Pattern-based	✅ Pattern-based	✅ With checksum
FEAT policy templates (credit, trading, insurance)	❌	❌	✅ Full library
AI System Registry with runtime enforcement	❌	❌	✅
FEAT assessment workflows (4 pillars)	❌	❌	✅
Kill switch (runtime enforcement)	❌	❌	✅
Bias monitoring with threshold alerts	❌	❌	✅
MAS-compliant export format	❌	❌	✅
7-year audit retention	❌	❌	✅
EU Banking Compliance (EBA ML Guidelines)
EU Banking PII detection (IBAN, BIC, VAT)	❌	❌	✅ With MOD-97
EBA ML policy templates (explainability, IRB)	❌	❌	✅ Full library
Model Registry with runtime enforcement	❌	❌	✅
Model validation workflows	❌	❌	✅
Explainability logging (Shapley values)	❌	❌	✅
Model change management (CA approval)	❌	❌	✅
EBA-compliant export format	❌	❌	✅

3. Policy Management

Policy hierarchy, versioning, and configuration capabilities.

Feature	Community	Evaluation	Enterprise
63 System policies (SQLi, PII, compliance)	✅ View	✅ View	✅ View + Override
Tenant-tier policies	✅ 20 limit	✅ 50 limit	✅ Unlimited
Organization-tier policies	❌	✅ 5 limit	✅ Full CRUD
System policy overrides (block→warn→log)	❌	❌	✅
Policy version history	✅ Last 5	✅ Last 5	✅ Full audit trail
require_approval action	✅ Auto-approves	✅ Auto-approves	✅ Full HITL queue
Policy templates library	Core	Core	Full (EU AI Act, HIPAA, PCI-DSS, SEBI, RBI, MAS FEAT, EBA ML)
Customer Portal Policy UI	❌	❌	✅

4. LLM Providers & Routing

LLM provider support, routing strategies, and cost management.

Feature	Community	Evaluation	Enterprise
Providers
OpenAI	✅	✅	✅
Azure OpenAI	✅	✅	✅
Anthropic (Claude)	✅	✅	✅
Google Gemini	✅	✅	✅
Ollama	✅	✅	✅
AWS Bedrock	❌	❌	✅
Routing
Multi-provider failover	✅	✅	✅
Weighted routing strategy	✅	✅	✅
Round-robin routing	✅	✅	✅
Failover-only strategy	✅	✅	✅
Health-based selection	✅	✅	✅
Cost-optimized routing	❌	❌	✅
Runtime weight updates	❌	❌	✅
Customer Portal provider UI	❌	❌	✅
Cost Management
Usage tracking (tokens, cost)	✅	✅	✅
Budget limits (org, team, agent, workflow)	✅	✅	✅
Budget alerts and thresholds	✅	✅	✅
Pre-flight budget check	✅	✅	✅
Usage breakdown by dimension	✅	✅	✅
Usage forecast	❌	❌	✅
Usage export (billing integration)	❌	❌	✅
Alert channels (Slack, email, webhook)	❌	❌	✅
Auto-downgrade to cheaper models	❌	❌	✅
Budget rollover	❌	❌	✅
Cost dashboard	❌	❌	✅
Cost estimation	Aggregate only, 10/day	Full breakdown, 100/day	Unlimited

5. MCP Connectors & Data Access

MCP connector availability and policy enforcement for data access.

Feature	Community	Evaluation	Enterprise
Available Connectors
PostgreSQL, MySQL, MongoDB	✅	✅	✅
Redis, HTTP/REST, Cassandra	✅	✅	✅
S3, Azure Blob, GCS	✅	✅	✅
Amadeus (Travel API)	❌	❌	✅
Salesforce	❌	❌	✅
Slack	❌	❌	✅
Snowflake	❌	❌	✅
HubSpot	❌	❌	✅
Jira	❌	❌	✅
ServiceNow	❌	❌	✅
Customer Portal Connector UI	❌	❌	✅
MCP Policy Enforcement
SQL injection blocking (REQUEST phase)	✅	✅	✅
PII redaction (RESPONSE phase)	✅	✅	✅
Exfiltration detection (row/volume limits)	✅	✅	✅
MCP query audit logging	✅	✅	✅
Tenant policy evaluation	✅	✅	✅
Rate limiting	✅	✅	✅
Budget controls	✅	✅	✅
Connectors with custom policies	2	5	Unlimited
Resource Limits
Tenant policies	20	50	Unlimited
Organization policies	0	5	Unlimited
LLM providers	2	3	Unlimited
Audit log retention	3 days	14 days	10 years
MAP plans	25	100	Unlimited
Versions per plan	10	25	Unlimited
Execution history	50	500	Unlimited
Concurrent executions	5	25	Unlimited
Time-window aggregation (exfiltration)	❌	❌	✅
ML-based anomaly detection	❌	❌	✅
Per-tenant limit overrides	❌	❌	✅
Cross-tenant policy inheritance	❌	❌	✅
Custom policy scripting	❌	❌	✅

6. Platform & Operations

Deployment, multi-agent planning, workflow control plane, and operational capabilities.

Feature	Community	Evaluation	Enterprise
Deployment
Docker Compose (local)	✅	✅	✅
AWS ECS/Fargate	Manual	Manual	One-click CloudFormation
Multi-tenant isolation	❌	❌	✅
Multi-Agent Planning (MAP)
YAML agent configuration	✅	✅	✅
Parallel task execution	✅	✅	✅
Conditional logic & branching	✅	✅	✅
Agent registry with hot reload	✅	✅	✅
REST API (list, get, validate)	✅	✅	✅
Decision & Execution Replay API	✅	✅	✅
axonctl CLI execution commands	✅	✅	✅
Embedded Execution Viewer (Web UI)	✅	✅	✅
REST API (CRUD, versions, sandbox)	❌	❌	✅
Database-backed agent storage	❌	❌	✅
Execution analytics dashboard	❌	❌	✅
Compliance export (PDF/CSV)	❌	❌	✅
Workflow Control Plane
Workflow registration (LangChain, LangGraph, CrewAI)	✅	✅	✅
Step gates (allow/block)	✅	✅	✅
Policy evaluation at step level	✅	✅	✅
SDK support (Go, Python, TypeScript, Java)	✅	✅	✅
LangGraph Python adapter	✅	✅	✅
require_approval action	✅ Returns decision	✅ Returns decision	✅ Routes to Portal HITL
failWorkflow()	✅	✅	✅
Pending execution approvals	Max 5 concurrent	Max 25 concurrent	Unlimited
HITL Queue API	❌	❌	✅
HITL expiration	❌	❌	✅ Auto-expiry
MAP-HITL integration	❌	❌	✅
Org-level workflow policies	❌	❌	✅
Cross-workflow analytics	❌	❌	✅
Step budgets/rate limits	❌	❌	✅
Support & Commercial
Community support (GitHub Issues)	✅	✅	✅
Priority support & SLA	❌	❌	✅
Customer dashboard UI	❌	❌	✅
Usage analytics & reporting	❌	❌	✅
AWS Marketplace integration	❌	❌	✅

---

When to Use Each Tier

Community Tier

Best for:

• Evaluation and POCs — Try AxonFlow locally before committing
• Development environments — Build and test with the same platform you'll use in production
• Small-scale deployments — Single teams with basic governance needs
• Learning — Understand how AI governance works hands-on

Get started:

git clone https://github.com/getaxonflow/axonflow.git
cd axonflow
docker compose up -d

Evaluation Tier

Best for:

• Serious evaluation — Need more policies and execution capacity than Community allows
• Pre-production validation — Organization-tier policies let you share governance rules across tenants
• Policy testing — Simulate policy changes and generate impact reports before deploying
• Human oversight evaluation — Test HITL approval gates with up to 100 pending requests
• Evidence gathering — Export watermarked evidence packs for internal reviews
• Time-bound evaluation — 14-day audit retention covers evaluation review cycles
• Growing evaluations — When 20 tenant policies or 5 concurrent executions aren't enough

Get Evaluation (free):

1. Register at getaxonflow.com/evaluation-license
2. Set the license key: export AXONFLOW_LICENSE_KEY=AXON-...
3. Restart AxonFlow — elevated limits are applied automatically

Enterprise Tier

Required when:

• Regulated industries — Healthcare (HIPAA), Finance (PCI-DSS, SEBI), EU (AI Act)
• Production workloads — High availability, auto-scaling, monitoring
• IT approval needed — SSO, SCIM, audit requirements
• Complex integrations — Amadeus, Salesforce, Slack, Snowflake, HubSpot, Jira, ServiceNow

Get Enterprise:

• AWS Marketplace: One-click deployment
• Contact Sales: [email protected]

---

Upgrade Path

The upgrade path is seamless at each step:

Community → Evaluation

1. Register for a free license at getaxonflow.com/evaluation-license
2. Set AXONFLOW_LICENSE_KEY in your environment
3. Restart AxonFlow — your existing policies, configurations, and data are preserved
4. Limits increase immediately (20→50 tenant policies, 0→5 org policies, 3→14 day retention, 50→500 execution history, 5→25 concurrent executions) and new capabilities unlock (HITL approval gates, policy simulation, impact reports, evidence export)

Evaluation → Enterprise

1. Contact [email protected] or deploy via AWS Marketplace
2. Replace your Evaluation license key with your Enterprise key
3. Enterprise features (SSO, SCIM, compliance modules, HITL queue) are enabled immediately
4. All limits become unlimited

Technical Details

• Same codebase — Each tier builds on the previous, so your policies and configurations carry over
• Same APIs — SDKs work identically across all tiers
• No migration needed — Just update your license key

SDK Version: Evaluation tier features (organization policies, tier-specific limits) require AxonFlow SDK v3.2.0 or later across all languages (Go, Python, TypeScript, Java).

Enterprise customers receive upgrade-safe guarantees. Community deployments that modify enforcement logic are unsupported and may diverge from compliance semantics over time.

Note: The license key (AXONFLOW_LICENSE_KEY) is configured as an environment variable on the AxonFlow server, not in the SDK client. The SDK connects to the server using clientId and clientSecret.

// Community — connect to local AxonFlow
const axonflow = new AxonFlow({
  endpoint: process.env.AXONFLOW_ENDPOINT || 'http://localhost:8080',
  clientId: process.env.AXONFLOW_CLIENT_ID || 'my-org',
  clientSecret: process.env.AXONFLOW_CLIENT_SECRET || 'my-secret'
});

// Evaluation — same SDK config; the Evaluation license key is set on the server
const axonflow = new AxonFlow({
  endpoint: process.env.AXONFLOW_ENDPOINT || 'http://localhost:8080',
  clientId: process.env.AXONFLOW_CLIENT_ID || 'my-org',
  clientSecret: process.env.AXONFLOW_CLIENT_SECRET || 'my-secret'
});

// Enterprise — point to your production endpoint
const axonflow = new AxonFlow({
  endpoint: process.env.AXONFLOW_ENDPOINT || 'https://your-agent.aws.com',
  clientId: process.env.AXONFLOW_CLIENT_ID || 'my-org',
  clientSecret: process.env.AXONFLOW_CLIENT_SECRET || 'my-secret'
});

---

Example Demos

Demo	Tier	Features Used
Getting Started	Community	PostgreSQL, OpenAI
Org Policies	Evaluation	Organization-tier policies
Healthcare Example	Enterprise	Bedrock, HIPAA policies
Trip Planner	Enterprise	Amadeus, Bedrock, MAP
E-commerce	Enterprise	Salesforce, PII detection

Footnotes

1. Measured on policy-only paths; orchestration latency varies by workflow complexity. ↩