TCP/IP (Transmission Control Protocol/Internet Protocol) is the foundational communication protocol suite that powers the internet and most computer networks. It provides reliable, ordered delivery of data between applications across diverse network hardware through a layered architecture of protocols.
TCP/IP
TCP/IP, the Transmission Control Protocol/Internet Protocol suite, is the layered set of protocols that moves data across the internet and nearly every computer network beneath it. It is the plumbing every API quietly depends on: IP routes packets between hosts, and TCP turns that best-effort delivery into the reliable, ordered byte stream that HTTP, TLS, and ultimately your API calls ride on top of.
- Layered architecture - The suite separates concerns into link, internet, transport, and application layers so each can evolve independently.
- Reliable transport (TCP) - Connection setup, sequencing, retransmission, and flow control guarantee ordered, lossless delivery.
- Addressing and routing (IP) - Every host gets an address, and packets find their way across arbitrary network hardware.
- Foundation for the web - HTTP, TLS, DNS, and the rest of the API stack are all built directly on TCP/IP.
In real API operations TCP/IP is rarely something teams touch directly, but it shapes everything above it — connection reuse, keep-alive, latency budgets, and timeouts all trace back to how TCP behaves. As the industry moves toward QUIC and HTTP/3, some of TCP’s role shifts onto UDP, but the addressing and layering model of the suite remains the mental map for reasoning about how requests actually travel. Understanding it is what separates surface-level API design from engineers who can debug why a call is slow, dropped, or stalled at the wire.
Referenced on the API Evangelist blog
Where this standard shows up across sixteen years of my writing at apievangelist.com — how it fits into API design, governance, and the agentic turn.