AI Coding Agents

AWS's spec-driven agentic dev environment: a prompt becomes requirements, a design and sequenced tasks.

Best for: Large existing codebases where a chat assistant loses the thread

OpenAI's coding agent for the terminal, IDE, desktop and cloud sandboxes, included with ChatGPT plans or billed by API tokens.

Best for: ChatGPT subscribers who want a coding agent included in their plan

Anthropic's coding agent that reads a repository, edits files and runs commands from the terminal, IDE, desktop app, browser or Slack

Best for: Developers who live in the terminal and want an agent that runs tests and opens PRs rather than suggesting lines

Browser workspace where an agent builds and ships apps

Best for: Solo builders who want code, database and hosting in one account without local setup

Free, open-source coding agent built on Rust and Tauri, with bring-your-own-key access to OpenAI, Anthropic, Google or local models.

Best for: Developers who already pay for Claude Pro, ChatGPT Plus or Copilot and want to reuse it

Coding agent running several tasks in parallel from a Mac app, a VS Code or JetBrains extension, or Slack

Best for: Developers who want several tasks running against one repo at once

AI agent that works inside the Unity Editor: writes scripts, builds scenes, debugs in play mode.

Best for: Unity teams wanting an assistant that understands scenes and prefabs, not just C# files

Autonomous software engineer working in a cloud VM

Best for: Engineering teams delegating well-specified tickets such as migrations, test coverage and dependency upgrades

Open-source AI coding agent and autocomplete that runs in your IDE, in the cloud or fully self-hosted

Best for: Teams whose code must stay on-prem but still want an agent and completions

Open-source terminal pair programmer that edits a git repository with the model you choose and commits each change.

Best for: Developers who want model choice and bring-your-own-key billing rather than a bundled subscription

Terminal-based coding agent that edits your codebase and runs commands from natural-language requests.

Best for: Developers who prefer terminal workflows over an IDE plugin

AI teammates that write code, browse the web, and run scheduled tasks for a startup.

Best for: First-time founders who want automated coding, research and scheduled tasks in one place

Inline completions, chat and coding agents across the major IDEs

Best for: Teams standardising on one assistant across mixed IDEs and languages

Open-source coding agent for the terminal, desktop and IDE that works with 75+ model providers or your existing Copilot and ChatGPT plans.

Best for: Developers who want to switch freely between Claude, GPT, Gemini and open-weight models

GitHub-connected coding agent that builds full-stack apps, works existing repos and reviews pull requests

Best for: Teams that want an agent working in GitHub pull requests rather than inside an editor

AI coding agent for terminal, cloud and desktop, backed by its own coding models and air-gapped deployments for regulated teams.

Best for: Banks, defence and healthcare engineering teams that need an on-premises or air-gapped coding agent

Google's agent-first coding platform: a desktop agent manager, terminal CLI, Python SDK and IDE built around Gemini models.

Best for: Former Gemini CLI users who need a supported terminal agent

Sawyer, an agent that lives inside Unity and takes on bug fixes, scene edits and SDK integration.

Best for: Unity studios with a backlog of engine-level bugs

Desktop coding agent from Z.ai that runs multi-agent, long-running tasks against the GLM-5.3 model.

Best for: Developers already committed to the GLM model family who want a native desktop agent

Cloud coding agent, formerly Ara, that turns assigned GitHub issues into pull requests with recordings, logs and diffs.

Best for: Engineering teams that triage work through GitHub issues

AI coding environment from ByteDance: the TraeCode IDE with a SOLO agent mode, plus TraeWork, a cloud agent for everyday tasks.

Best for: Developers trying an agent-first IDE with a free starting tier