Cline
Cline is an open-source coding agent under Apache-2.0 from Cline Bot Inc. that runs in VS Code, JetBrains IDEs, a terminal and a desktop app. It makes coordinated edits across a codebase as diffs you review, runs terminal commands and watches the output, and requires your approval for each action unless you turn that off. It works with any model provider, including local models through Ollama, and connects tools through MCP.
What does Cline do?
Cline calls itself "the open source coding agent in your IDE, terminal, & desktop." It edits code and runs commands as an agent rather than a completion engine: the README says it "makes coordinated changes across your codebase", that "every edit shows up as a diff you can review, modify, or revert", that changes are tracked with checkpoints, and that it "executes commands directly in your terminal and watches the output in real time." It has a Plan mode and an Act mode, and "every file edit and terminal command requires your approval", with optional auto-approve.
It ships as a VS Code extension, a JetBrains plugin for IntelliJ IDEA, PyCharm, WebStorm, GoLand and the rest of that family, a CLI that runs interactively or fully headless for CI and scripting with JSON output, a native desktop app for macOS and Windows, and a Node.js SDK. The README lists providers: Anthropic, OpenAI, Google, OpenRouter with over 200 models, Vercel AI Gateway, AWS Bedrock, Azure and GCP Vertex, Cerebras and Groq, Ollama and LM Studio for local models, and any OpenAI-compatible API. MCP servers connect databases, APIs and cloud infrastructure, managed in the CLI with cline mcp. Newer features named: multi-agent teams, scheduled agents with cron, and connectors for Telegram, Slack, Discord, Google Chat, WhatsApp and Linear.
Key facts
- Licence: Apache-2.0, copyright Cline Bot Inc., with no added conditions. The README notes one exception: "Currently we are not open-sourcing JetBrains plugins."
- Surfaces: VS Code extension, JetBrains plugin, CLI (interactive or headless with JSON output), desktop app for macOS and Windows, and the @cline/sdk Node.js package.
- Every file edit and terminal command requires approval by default; auto-approve is optional. Edits appear as diffs with checkpoints to revert.
- Providers: Anthropic, OpenAI, Google, OpenRouter (200+ models), Vercel AI Gateway, AWS Bedrock, Azure and GCP Vertex, Cerebras, Groq, Ollama, LM Studio, and any OpenAI-compatible API.
- MCP support for tools; .clinerules files for project instructions; plugins through the SDK.
- Multi-agent teams with cline --team-name, scheduled agents with cline schedule create and a cron expression, and chat connectors for Telegram, Slack, Discord, Google Chat, WhatsApp and Linear.
- Install: npm i -g cline for the CLI, the VS Code and JetBrains marketplaces, or the desktop download at cline.bot/desktop.
- Latest tags when read: v4.1.19 on 2026-09-17 and desktop-v0.0.31 the same day; 68,561 stars on 2026-09-18.
What does it replace, and where does it fall short?
Cline replaces the agent half of Cursor, Windsurf and GitHub Copilot inside the editor you already use, and its CLI replaces Claude Code in a terminal, with the model chosen by you and the diffs reviewed before they land. Its README makes the point directly: "Cline is not locked to a single AI provider." On the open-source alternatives to Cursor page it is the pick for staying in VS Code or JetBrains with any model.
Where it falls short: the README does not describe inline autocomplete, so the ghost-text completion that Copilot and Cursor lead with is not part of Cline; Kilo Code and Tabby cover that. The JetBrains plugin is not open source. Every provider is its own account and bill.
How does Cline run?
For the CLI, npm i -g cline. For VS Code, install from the marketplace; for JetBrains, from the JetBrains Marketplace; the desktop app downloads for macOS and Windows; the SDK installs with npm install @cline/sdk. It needs a model provider of your choice: an API key for a hosted one, or Ollama or LM Studio for a local model. The README does not describe a hosted or paid Cline tier; the marketplaces and the desktop app are distribution.
Who is Cline for?
A developer who wants an agent that edits and runs things under review, in the IDE they already have, with the freedom to move between models and to run the same agent headless in CI. Someone whose main want is fast inline completion should look at Tabby, twinny or Kilo Code.
What limits does the README state?
From the README: "Currently we are not open-sourcing JetBrains plugins", and the VS Code extension location is marked as a work in progress while it migrates.
Questions people ask
Is Cline open source?
Yes. The Cline repository is Apache-2.0 with no added conditions, which covers the CLI, the VS Code extension, the desktop app and the SDK. The one exception the README states is the JetBrains plugin, which is not open-sourced at present.
Can Cline use local models?
Yes. The README lists Ollama and LM Studio under "Run local models on your machine", alongside any OpenAI-compatible API. Hosted providers listed include Anthropic, OpenAI, Google, OpenRouter, Bedrock, Azure, Vertex, Cerebras and Groq, and you can switch between them.
Does Cline run commands without asking?
Not by default. The README says every file edit and terminal command requires your approval, so you stay in control of what changes. Auto-approve is an option you turn on, and the headless CLI mode runs with zero interaction for scripting and CI.
How does Cline compare with Cursor and Claude Code?
Cline gives you the agent inside VS Code or JetBrains with any model, or in the terminal, with diffs to review and no platform subscription. Cursor is an editor with its own completion and agent; Claude Code is Anthropic's terminal agent for Claude models. See the alternatives to Cursor and alternatives to Claude Code pages.
Does Cline support MCP?
Yes. The README describes using MCPs to connect to databases, query APIs and manage cloud infrastructure, and the CLI manages servers with cline mcp. Beyond MCP it offers plugins through the SDK, .clinerules project files, and connectors for Slack, Telegram, Discord, Google Chat, WhatsApp and Linear.
Sources
- Cline README and Apache-2.0 LICENSE: github.com/cline/cline, read 2026-09-17.
Compared with the others
On the open-source alternatives to Cursor page, Cline is the pick for an agent with reviewable diffs and any model. Apache-2.0, in VS Code, JetBrains, the terminal and a desktop app, with every edit shown as a diff, approval on each action, Ollama and LM Studio for local models, and MCP.
Also on that page: Zed for replacing the editor itself, Kilo Code for autocomplete plus agent, inside vs code or jetbrains, Tabby for a completion server for a whole team.
On the open-source alternatives to GitHub Copilot page, Cline is the pick for the chat and agent side, any model. Apache-2.0, with reviewable diffs, approval on every action, MCP, and Ollama or LM Studio for local models.
Also on that page: Tabby for a self-hosted copilot server for a team, twinny for ghost text in vs code from a local model, Kilo Code for completion and an agent in one extension.
On the open-source alternatives to Windsurf page, Cline is the pick for an agent with reviewable diffs. Apache-2.0, any provider including Ollama and LM Studio, approval on every edit and command, MCP, plus a CLI and a desktop app.
Also on that page: Zed for an open editor that hosts other agents, Kilo Code for autocomplete plus agent in vs code or jetbrains, Tabby for team completion on your own gpu.
On the open-source alternatives to Claude Code page, Cline is the pick for the same agent in the ide and headless in ci. Apache-2.0, a CLI with JSON output for scripting, VS Code and JetBrains extensions, a desktop app, any provider including Ollama, and approval on every action.
Also on that page: OpenCode for a terminal agent with a planning mode, Qwen Code for switching between hosted and local models at runtime, Crush for the widest provider list with local models.
On the open-source alternatives to Devin page, Cline is the pick for headless runs in ci and scheduled agents. Apache-2.0, a CLI that runs with zero interaction and JSON output, scheduled agents with cron, multi-agent teams, and connectors for Slack, Telegram, Discord and Linear.
Also on that page: OpenHands for an always-on agent on your own infrastructure, goose for a general agent with a desktop app and an api, Kilo Code for a task run from the web.
More coding assistants and agents
An MIT terminal coding agent with a full-access build agent and a read-only plan agent switched with the Tab key, installers for every platform, a beta desktop app, and the most-starred repository in this directory.
OpenAI's Apache-2.0 coding agent that runs locally in your terminal, with a desktop app and IDE extensions, signed in with a ChatGPT plan or an OpenAI API key, so the code is open and the models are OpenAI's.
Google's Apache-2.0 terminal agent for Gemini models with a free tier of 1,000 requests a day on a personal Google account, built-in search grounding, file and shell tools, MCP support, checkpointing and a GitHub Action.
A high-performance code editor from the creators of Atom and Tree-sitter, GPL-3.0-or-later with Apache-2.0 components, with an agent panel, external agents over ACP, edit prediction, MCP tools, sandboxing, and local models through Ollama or llama.cpp.
An MIT self-hosted control centre, called Agent Canvas, that runs the OpenHands agent or Claude Code, Codex, Gemini and any ACP-compatible agent across local, Docker and cloud backends, with automations triggered by schedules and webhooks.
An Apache-2.0 general-purpose AI agent from the Agentic AI Foundation at the Linux Foundation, written in Rust, with a desktop app, a CLI and an API, 15+ model providers including Ollama, and 70+ extensions over MCP.
Added September 17, 2026. Every claim above comes from the project's README, LICENSE or model card, read on September 17, 2026, or from the GitHub API on the date shown in the panel. Found an error? Write to reveneau@licheo.com and it is fixed in the next weekly pass. Repository: github.com/cline/cline.
Running one of these inside your own environment, with your own data and your own security rules, is the kind of work Reveneau does. Read how a forward deployed engagement works.