Agentic coding

Agentic coding means a model works on software like a developer: it reads the codebase, edits files, runs commands and tests, and keeps going until the task is done.

  • Using models
  • Updated

A plain code assistant completes the line you are typing. A coding agent receives a task, such as fixing a bug, and works in a loop: it explores the repository, makes changes, runs the tests, reads the errors and tries again, often over dozens or hundreds of tool calls. This needs a long context, reliable tool calling and the stamina not to lose the thread.

Developers describe such models with benchmarks like SWE-bench and Terminal-Bench. Scores in model announcements are self-reported and depend on the agent harness around the model. Try two or three models in your own harness, on your own repository, before you commit to one.

Related terms

  • AI agent An AI agent is a system in which a model plans and carries out multi-step tasks by calling tools, such as running code, searching the web or editing files.
  • Tool calling (function calling) Tool calling lets a model ask your program to run a function, such as a search or a calculator, and then use the result in its answer.
  • Context window The context window is the maximum amount of text, measured in tokens, that a model can take into account at one time, including your prompt and its own answer.
  • Reasoning model A reasoning model works through a problem step by step in a 'thinking' phase, visible or hidden, before it gives its final answer.

Keep reading

All terms