AI Gateway pricing
Pricing for governed LLM traffic
Metered on gateway-processed requests (not provider tokens). Model spend is billed by your providers; ForgeCrux meters routing, guardrails, cache, and observability.
Priced per million AI requests / month
Team
Copilots and internal apps
$2,000
/ month · up to 5M requests
- OpenAI-compatible endpoint
- Multi-model routing + fallback
- Prompt registry
- PII + injection guardrails
- Token and spend dashboards
- Email support
Most chosen
Enterprise
Company-wide AI control plane
$9,000
/ month · 25M requests included
- Cost/latency/quality routing
- Semantic cache
- Eval gates + online quality
- Residency and provider allow lists
- VPC / hybrid runtime
- 99.99% SLA + 24×7 P1
Regulated
Sovereign and air-gapped AI
Custom
annual license + capacity
- Self-hosted model pools only
- Air-gapped policy packs
- Customer-managed keys
- Long-term WORM audit
- On-prem eval cluster
- Dedicated security reviews
- • Provider invoices (OpenAI, Anthropic, Bedrock, etc.) remain with you.
- • Cached hits count at 20% of a billable request on Enterprise and Regulated.
Product details: AI Gateway platform page