Description
F5 AI Gateway is F5's product for governing how an organization uses AI: every model call, every AI agent, and every tool an agent might invoke, routed through one control plane instead of scattered integrations. F5 positions it as covering three things at once, a Model Gateway, an MCP Gateway for agent-to-tool traffic using the Model Context Protocol, and AI Guardrails, under a single policy model, and states it requires no changes to the applications that call it.
Read More
On cost, the product meters tokens and enforces budgets by provider, model, team, and individual user, giving full visibility into where AI spend goes and letting teams cap it. F5 states this combination of semantic caching, smart routing to the most cost-effective model for a given task, and an MCP layer that removes redundant agentic tool calls can cut token costs by 30 to 60 percent, and that MCP-level deduplication alone can cut token waste from repeated tool calls by up to 90 percent.
On security, AI Gateway inspects prompts and responses in real time to redact sensitive data such as source code, credentials, or customer information before it reaches an external model, and to block prompt-injection and jailbreak attempts with fail-closed enforcement, backed by what F5 describes as a threat library updated with over 10,000 new attack patterns a month. It maintains audit trails aligned to SOC2, ISO, and HIPAA, with SIEM export and data-residency controls. On governance, an MCP Server Registry gives a single view of every public, remote, and private MCP server in use, with per-tool allow and deny lists, access controls, spending limits, and a full audit trail of every tool call.
F5 AI Gateway is vendor-agnostic, routing traffic across multiple AI providers and models from one control point, and is aimed at enterprise security, platform, and AI operations teams managing AI usage across an organization rather than individual developers. The product page doesn't list public pricing; it directs visitors to sign up for access rather than a self-serve price list.
- CategoryAI DevOps & Cloud Tools
- Pricing ModelSubscription
- DeploymentMulticloud
- PlatformsCloud
F5 AI Gateway's Core Features
Model Gateway that routes, caches, and optimizes LLM requests across providers
MCP Gateway governing agent-to-tool calls under the Model Context Protocol
Per-provider, per-model, per-team, and per-user token metering and budgets
Semantic and exact caching plus GPU-aware routing to cut cost and latency
20+ AI Guardrail policies for PII/PHI redaction and prompt-injection defense
MCP Server Registry listing every public, remote, and private MCP server in use
Per-tool access controls, allow/deny lists, and spending limits
Full audit trails aligned to SOC2, ISO, and HIPAA with SIEM export
Vendor-agnostic routing across multiple AI providers and models
How to use F5 AI Gateway?
Sign up for access: Request access to F5 AI Gateway.
Register models and MCP servers: Add the AI models and MCP tools the organization uses to the asset registry.
Set budgets and guardrails: Configure token budgets per team and enable guardrail policies for data and safety.
Route traffic through the gateway: Point applications and agents at the Model Gateway and MCP Gateway.
Monitor and audit: Use the dashboards and audit trail to track spend, blocked threats, and tool calls.
F5 AI Gateway's Use Cases
- AI cost governance
- Sensitive data protection
- AI agent and MCP governance
- Regulated AI deployments
