AI Security & Expert Personas

Secure AI — for your gateway and your team

Gateway security for every AI call — plus expert AI personas grounded in your org's documents, with managed AI included on every tier. No API key needed to get started. Three tiers, published prices. Start with a 7-day free trial on Starter — no credit card.

Monthly Annual 2 months free
🔑

Fire Mission charges for the security layer — not for AI inference. You pay OpenAI, Anthropic, Google, Groq, and other providers directly at their published rates. Your token costs never pass through our billing. Fire Mission adds zero markup on inference. Expert personas use platform-funded managed AI — no provider key needed; monthly query credits are included per tier below.

Starter

$19 /seat/mo

7-day free trial · then /seat/mo

7-day free trial — 1 AI Persona, Standard & Advanced AI models, 100 managed AI queries/month

  • 1 AI Persona (from curated catalog or custom-built) — published to your team at /app
  • 100 managed AI queries/month — Standard (Groq Llama 70B) & Advanced (GPT-4o Mini / OpenRouter) AI
  • Active AI DLP blocking — violations stopped in-flight, not just logged
  • Mandatory security scanning: PII, prompt injection, model substitution detection
  • 30-day audit log retention
  • Add Gold AI ($29/mo) to unlock Reasoning-class models (Claude Sonnet)
Start 7-day trial
Most popular

Professional

$49 /seat/mo

billed monthly

3 AI Personas, Professional AI (GPT-4o + Gemini Flash), 300 managed queries/month, audit export

  • 3 AI Personas (unlimited custom-built from catalog) — published to your team at /app
  • 300 managed AI queries/month — Professional AI: GPT-4o, Gemini 2.5 Flash, OpenRouter
  • Active AI DLP blocking + custom policy enforcement
  • 4 BYOK provider keys (admin opt-in required)
  • 90-day audit log retention with CSV/NDJSON export
  • Executive AI Risk Dashboard + SIEM webhook (Splunk, Elastic, Datadog)
  • Add Gold AI ($29/mo) to unlock Reasoning-class models (Claude Sonnet)
Get Professional

Enterprise

$149 /seat/mo

billed monthly

10 AI Personas, Quality AI (Claude Haiku + Gemini Pro), 1,000 managed queries/month, SCIM & SLA

  • 10 AI Personas — unlimited custom-built — 20 docs/persona — published to your team at /app
  • 1,000 managed AI queries/month — Quality AI: Claude Haiku, Gemini 2.5 Pro, GPT-4o
  • Everything in Professional, plus BYOK support (admin opt-in), SCIM, on-prem deployment, and SLA
  • SCIM 2.0 user provisioning (Okta, Azure AD, Google Workspace)
  • Unlimited audit log retention + Uptime SLA + dedicated security engineer
  • Add Gold AI ($29/mo) to unlock Reasoning-class models (Claude Sonnet)
Get Enterprise

Gold AI Add-on

Optional upgrade

Unlock Reasoning-class AI models (Claude Sonnet) on any Starter, Professional, or Enterprise plan. Gold queries cost 3 credits each and produce deeper, step-by-step analysis for complex decisions.

  • ✓ Claude Sonnet via OpenRouter — Reasoning-class inference
  • ✓ Available on all tiers — add any time from your dashboard
  • ✓ $29/month flat — no per-seat charges
$29
/org/mo
Get started

Full feature comparison

Gateway security rows are enforced in middleware — not a policy checkbox. Persona and managed AI rows reflect platform capabilities available on every tier.

Included Not available on this tier ✅ Active block Active enforcement
Feature
Starter
$19/seat
Professional
$49/seat
Enterprise
$149/seat
🛡️Gateway Security
BYOK proxy (bring your own AI provider key)
Mandatory security scan on every call
Prompt injection & jailbreak blocking
API key leakage detection
LLM model substitution detection
Monthly gateway call cap UnlimitedUnlimitedUnlimited
Passthrough Monitor (log-only lane) ✅ All tiers✅ All tiers✅ All tiers
BYOK provider keys included 2 keys4 keysUnlimited
🔒AI Data Protection
AI DLP enforcement mode ✅ Active block✅ Active block✅ Active block
Custom security policies
📋Compliance & Audit
Audit log retention 30 days90 days365 days
Audit log export (CSV / NDJSON)
SIEM webhook push
Executive AI Risk Dashboard
🤖Team AI Portal
Total personas per org Up to 1Up to 3Up to 10
Custom personas (built from scratch) 1 (rest from catalog)UnlimitedUnlimited
Document sources per persona 5 per persona10 per persona20 per persona
System prompt length 1,000 chars3,000 chars8,000 chars
Round table simultaneous personas Up to 3Up to 3All personas
AI Personas portal — publish to team at /app
Managed AI queries / month 100 credits300 credits1,000 credits
Unlimited queries with BYOK ✅ All tiers✅ All tiers✅ All tiers
Ask an Expert (1 credit) ✅ All tiers✅ All tiers✅ All tiers
Team Briefing — parallel swarm (N+1 credits)
Cross-Check — consensus verification (3 credits) ✅ All tiers✅ All tiers✅ All tiers
🏢Enterprise Identity
SCIM 2.0 user provisioning
Deployment & SLA
On-premises / air-gapped deployment
Uptime SLA + priority incident response
Support EmailPriorityPriority + dedicated engineer

All tiers enforce active DLP blocking — PII violations are stopped in-flight before reaching the AI provider. Prompt injection and API key leakage are also hard-blocked on all tiers.

🔐

Zero prompt storage

Fire Mission logs token counts, cost, latency, and security verdicts — never raw prompt or completion bytes. Two-plane isolation: metadata only on the control plane.

🌐

US-only provider policy

Geopolitical supply chain enforcement built in. All approved AI providers must serve from US-based infrastructure. Foreign-jurisdiction providers are blocked at the gateway.

📜

Compliance posture

SOC 2-aligned architecture · NIST 800-53 aligned · HIPAA-ready architecture · FedRAMP-aligned · Stripe-powered checkout

Start your 7-day trial — active protection from day one

Full Starter tier access for 7 days: active DLP blocking, unlimited gateway calls, 30-day audit retention, and 100 managed AI queries — no API key needed. No credit card required to start.

Defense, federal, or DoW deployment? See Fire Mission Defense Edition →

All prices are USD and exclude applicable taxes. Per-seat pricing applies to named users. Seat minimums (Professional: 1, Enterprise: 1) are enforced at checkout.

Compliance posture: "SOC 2-aligned architecture" (not yet certified) · "HIPAA-ready" (not HIPAA compliant) · "FedRAMP-aligned" (not FedRAMP authorized) · "NIST 800-53 aligned" (not certified).