F5 adds token tracking

F5 upgraded its AI Gateway to report token usage by model, user, and team. This marks a major step into tokenomics, the management of AI costs.

Security meets cost control

The AI Gateway now integrates with F5's AI Security Platform. It provides real-time policy enforcement at the token level.

What AI gateways do

AI gateways act as middleware between applications and AI models. They provide security, load balancing, data orchestration, and inference cost management.

The fragmented AI governance problem

Enterprises are juggling multiple products to avoid security risks, cost overruns, and GPU lags. The current landscape is fragmented by design.

“Enterprises are stitching together bespoke proxies, provider-specific SDKs, standalone guardrails, and separate observability tools”

Competing in tokenomics

Other vendors like PortKey, Helicone, Langfuse, and Cloudflare AI Gateway offer similar token control. Tracking tokens is key to realizing ROI from AI projects.

F5's competitive advantage

F5's longstanding presence in application delivery and security, especially in multicloud environments, gives it an edge in this market.

Model Gateway routes smartly

Model Gateway abstracts LLM complexity and routes tasks to the best model. It caches prompts and tracks token spend by date, model, user, and team.

MCP Gateway controls agents

MCP Gateway shows every MCP server an organization uses. Access controls prevent agents from using unauthorized tools that could rack up costs.

“A single autonomous loop can consume thousands of dollars in tokens before anyone notices”

AI Guardrails protect data

F5 AI Guardrails tracks and redacts personally identifiable information and prompt injections. It enforces custom compliance and regulatory rules.

Tool sprawl will consolidate

As inference spreads in enterprises, the need for AI traffic controls grows. Products like F5's will address the current trend toward tool sprawl.