AI Gateway

An AI gateway is the control point that decides which provider, model, and API key serve each LLM request. This guide compares Bifrost, Kong AI Gateway, LiteLLM, Cloudflare AI Gateway, and Azure API Management on routing rules, load balancing, fallbacks, and governance of routes.

Deployment model decides most LLM gateway choices before features do. This guide compares Bifrost, Cloudflare, Kong, LiteLLM, and OpenRouter on deployment, published overhead, governance depth, and whether they govern agent tool calls.

An AI gateway manages Codex CLI token spend by enforcing budgets, caching repeated requests, and logging cost for every call. This guide covers how to check Codex CLI token usage, Bifrost's budget hierarchy, cost-aware routing, and the correct config.toml setup.

Codex security is the set of controls that decide what OpenAI Codex can reach, send, and change across an engineering organization. This guide compares six tools, including Bifrost, OpenAI managed configuration, Codex Security, and Amazon Bedrock Guardrails.

AI cost management is the practice of measuring, attributing, and controlling what an organization spends on LLM APIs. This guide compares five tools for tracking LLM costs across providers, including Bifrost, Datadog, Vantage, and cloud-native billing, on attribution depth, enforcement, and export.

Rate limiting in LLM apps is two problems: provider quotas you cannot change, and internal quotas you must enforce. This guide covers the limit dimensions, key pooling, and per-consumer virtual key limits at the gateway.