Search the
working landscape.

An open-source desktop workbench that runs coding agents in isolated Git worktrees and tracks each task through pull request, CI, review, and merge.

An open-source terminal pair programmer with repository maps, Git integration, and automated lint-and-test loops.

AWS’s IDE and CLI coding assistant, combining completion, agentic requests, transformation workflows, and enterprise controls.

A proprietary coding agent that runs each thread locally or in a per-thread remote machine (an orb), with web, CLI, and macOS/iOS clients and a meta-agent for coordinating parallel work.

Google’s multi-agent coding environment, combining a desktop editor and terminal CLI on one agent platform.

A team coding platform that runs event-triggered cloud agents (Cosmos) alongside the Auggie CLI, IDE extensions, and GitHub code review, all backed by cloud codebase indexing.

An agentic coding tool that reads, edits, and runs a codebase across terminal, IDE, desktop, and browser surfaces.

An open-source coding agent across VS Code, JetBrains, CLI, and a Kanban surface for parallel agent work.

OpenAI’s coding agent across ChatGPT, desktop, editor, terminal, worktrees, and cloud environments.

A discontinued open-source IDE and terminal coding agent retained to document its final release and migration context.

An AI-native desktop editor that combines completion, interactive agents, cloud agents, and review in one product.

Cognition’s local-and-cloud coding platform for delegating, supervising, reviewing, and shipping work across agent fleets.

Google’s open-source terminal coding agent, retained here as a deprecated predecessor to Antigravity CLI.

A multi-surface coding assistant spanning editor completion, chat, local agents, GitHub-native cloud agents, and review.

An open-source, provider-agnostic local agent with a CLI and desktop app, extended through MCP and reusable recipes.

Google’s asynchronous coding agent that plans and executes GitHub repository tasks in cloud VMs and returns branches for pull requests.

JetBrains’ coding agent for IDE and CLI workflows, with project guidelines, skills, MCP, subagents, and multiple model paths.

An open-source, self-hosted workbench for running multiple coding agents in parallel worktrees and reviewing their output.

AWS’s multi-surface coding environment, sharing specs, steering, hooks, skills, and one agent harness across IDE, CLI, and web.

A provider-independent, open-source coding agent for terminal, desktop, and editor workflows.

An open-source agent platform with Agent Canvas for local or remote backends, a coding SDK, managed cloud access, and enterprise deployment.

A discontinued open-source VS Code agent retained to document its shutdown and help existing users evaluate migration paths.

An enterprise AI coding platform combining completion, chat, agents, a context engine, and private deployment choices.

A modern terminal with local and cloud coding agents, multi-model access, team collaboration, and usage-based credits.

A fast, open-source desktop editor with native agents, edit prediction, provider choice, and parallel agent threads.

Browse tool records classified as ai-native editor.

Browse tool records classified as ide extension.

Browse tool records classified as terminal agent.

Browse tool records classified as cloud agent.

Browse tool records classified as orchestrator.

Browse tool records classified as open source.

Choose Cursor when the editor is the product, Claude Code when the terminal is the control plane, and Codex when local work and cloud delegation should share one OpenAI account.

Choose Claude Code when the repository and terminal should remain the stable control plane; choose Cursor when the editor should unify completion, interactive agents, review, and background work.

Choose Cursor when the team wants an AI-native editor as the primary environment; choose GitHub Copilot when developers should keep existing editors and center agent work on GitHub.

Choose Claude Code for a terminal-first agent with multiple access providers; choose Codex when local worktrees and OpenAI-managed cloud delegation should share one product family.

Choose Aider for a focused Git-aware terminal pair programmer; choose OpenCode when the same open-source agent should extend into desktop and editor surfaces.

Choose Cline for a maintained interactive editor agent. Keep Continue only for an existing final 2.0 installation while you plan migration; its repository is read-only.

Choose Cursor for an established completion-and-agent editor loop; choose Antigravity when Google-centered multi-agent supervision is more important than a mature editor ecosystem.

Choose OpenHands for an open local, cloud, or self-hosted path; choose Devin for a managed commercial platform designed around delegating and supervising agent fleets.

A practical guide to comparing AI coding-agent subscriptions, credits, provider usage, seats, and execution costs without pretending incompatible billing units are equal.

Choose an open-source or local AI coding agent by editor surface, provider flexibility, self-hosting, repository workflow, and the parts that still leave your machine.

Evaluate AI coding tools for a team through identity, model policy, data handling, execution isolation, usage controls, review gates, and migration cost.

The September 22 review clarifies Codex API-key billing, Tabnine inference charges, and the maintained OpenHands components to use in a new trial.

The continuedev/continue repository is read-only after a final 2.0.0 release of the VS Code extension, CLI, and JetBrains plugin, so the profile is recorded as discontinued and retained for migration context.

Amp, Augment Code, Jules, goose, and Agent Orchestrator join the landscape with primary-source evidence, filling the cloud-agent layer and adding a third orchestrator.

Claude Code’s terminal, editor, desktop, and web access were rechecked so the profile no longer frames the product as a CLI-only agent.

GitHub Copilot’s Free, Pro, Pro+, and Max individual plans and its agent-workflow AI-credit billing structure were rechecked against first-party pricing.

Gemini CLI was marked deprecated after its consumer request paths ended, with Antigravity CLI documented as Google’s active successor.

Kandev’s supported agents, worktree isolation, executors, task sources, review gates, and AGPL licensing were added to its evidence-backed profile.

Windsurf IDE is now represented inside Devin’s current local-and-cloud product and pricing structure rather than as a stale standalone profile.