52 facts per product, each taken from the vendor's own public pages and linked to its source. 13 rows differ and are highlighted; the better documented value in a row is bold. “Not documented” means the vendor's pages do not state it — it is never counted as a “No”.
Facts verified · Prices in USD · Machine-readable JSON
Cline installs as a free VS Code extension and has a JetBrains plugin in early access. Aider's pages point to third-party VS Code plugins and to running aider in an IDE's terminal, or to its watch mode, which acts on AI comments you type in any editor.
Aider sends a repo map with each request: the key classes, functions and signatures of the whole git repo, ranked by a file-dependency graph and trimmed to a token budget. Cline builds no index by design; it reads files and runs ripgrep searches as it goes.
Both can undo AI edits. Aider commits each of its edits to your git history, and /undo discards the last one. Cline keeps its checkpoints in a shadow git repository, separate from your project history.
Cline connects MCP servers and has an MCP marketplace for one-click installs. Aider's pages, including its options reference and release history, do not document MCP.
Cline's Enterprise plan, which is custom-priced, adds SSO, a team dashboard with per-team cost breakdowns, and VPC, on-prem or air-gapped deployment. Aider lists no team plan, admin controls or SSO.
Cline lists SOC 2 on its trust center, without a report type. Aider's pages document no certifications. Both run on your own machine, and both state they do not collect your code when you use your own API keys.
| Aider | Cline | |
|---|---|---|
| Free plan | Yes The tool is free and open source (Apache 2.0). Most models need an API key from an LLM provider, and you pay that provider for usage. | Yes The open-source client is free. Model usage is paid separately: Cline credits (pay as you go) or your own API key with a third-party provider. Local models and limited-time free-model promotions cost nothing. |
| Free plan limits | No aider-imposed limits; model usage is paid to your LLM provider (free options: OpenRouter free models with daily limits, local models) Aider never enforces token limits; it only reports limit errors from the API provider. The LLMs page names OpenRouter free models (limited daily usage) and Google's Gemini 2.5 Pro Exp as free options, and local models via Ollama. | Client free; AI inference paid separately (Cline credits or own API key); free models only as rotating limited-time promotions up to a usage quota Free-model usage is not available through the Cline API, and the docs say it "may be used to help improve model performance and quality". |
| Free trial (days) | Not documented Not documented on vendor pages | Not documented Not documented on vendor pages |
| Trial requires credit card | Not documented Not documented on vendor pages | Not documented Not documented on vendor pages |
| Cheapest paid plan, billed annually (per user / month) | Not documented Not documented on vendor pages | Not documented Not documented on vendor pages — ClinePass is listed only at $9.99/month; no annual billing is shown. |
| Cheapest paid plan, billed monthly (per user / month) | Not documented Not documented on vendor pages | $9.99 ClinePass: an optional subscription for curated open-weight models. The Cline client itself is free; the pricing page lists no other paid plan with a price. |
| Highest listed-price plan, billed annually (per user / month) | Not documented Not documented on vendor pages | Not documented Not documented on vendor pages — no plan lists an annual price; Enterprise is custom-priced. |
| Minimum paid seats | Not documented Not documented on vendor pages | Not documented Not documented on vendor pages |
| Custom-priced enterprise plan | Not documented Not documented on vendor pages | Yes Enterprise plan is priced "Custom" with a Contact Sales button. |
| Aider | Cline | |
|---|---|---|
| Web app | Yes Experimental browser UI that you launch locally with aider --browser; it edits files in your local git repo. Not a hosted web service. | Yes Cline Kanban (research preview) is a browser board served locally after `npx kanban` / `cline --kanban`; no vendor-hosted web IDE is documented. |
| Runs on Windows | Yes One-line PowerShell installer listed under the install page's WINDOWS heading. Also installs with pip, uv or pipx on Python 3.8-3.13. | Yes Via the CLI and editor extensions; Cline for Desktop also ships a Windows build (Windows support in beta). |
| Runs on macOS | Yes One-line curl/wget installer for Mac and Linux. Also installs with pip, uv or pipx on Python 3.8-3.13. | Yes Via the CLI and editor extensions; Cline for Desktop (beta) ships for macOS (Apple Silicon & Intel). |
| Runs on Linux | Yes One-line curl/wget installer for Mac and Linux. Docker images (paulgauthier/aider, paulgauthier/aider-full) are also published. | Yes Via the CLI and editor extensions; Cline for Desktop (beta) is listed for macOS and Windows only. |
| iOS app | Not documented Not documented on vendor pages | Not documented Not documented on vendor pages |
| Android app | Not documented Not documented on vendor pages | Not documented Not documented on vendor pages |
| Browser extension | Not documented Not documented on vendor pages | Not documented Not documented on vendor pages |
| Aider | Cline | |
|---|---|---|
| Own standalone code editor | Not documented Not documented on vendor pages | No Cline runs inside existing editors. Cline for Desktop (beta) is a standalone agent app for macOS and Windows, not a code editor. |
| Extension for VS Code | Not documented Not documented on vendor pages. The docs point to third-party aider plugins for VS Code and to running aider in VS Code's terminal. | Yes Free and open source; installed from the VS Code Marketplace (Open VSX for Windsurf and VSCodium). |
| Plugin for JetBrains IDEs | Not documented Not documented on vendor pages | Yes Labelled "Early Access". The pricing page lists "JetBrains Extension" under Enterprise features, while the IDE page and docs offer it to everyone. Unlike the rest of Cline, the JetBrains plugin is not open source. |
| Command-line (terminal) agent | Yes Aider is a terminal program; the browser UI and IDE file-watching are add-ons to it. | Yes Installed with `npm i -g cline`; interactive, one-shot and headless (CI/CD) modes. |
| Other editors supported | Any IDE or text editor via --watch-files AI comments; NeoVim via a community plugin (aider.nvim) With --watch-files, aider reads one-line comments ending in AI! or AI? that you type in any editor and acts on them. The NeoVim plugin is by a community member (joshuavial), not the vendor. | Cursor, Windsurf, VSCodium, Antigravity (extension); Zed, Neovim, Emacs and other ACP clients (via the CLI) The ACP docs also cover Emacs (agent-shell) and JetBrains AI Assistant, which can run Cline through the CLI with `cline --acp`. |
| Aider | Cline | |
|---|---|---|
| Inline code completion (autocomplete) | Not documented Not documented on vendor pages | Not documented Not documented on vendor pages |
| Agent that edits across multiple files | Yes Edits every file added to the chat and asks to add other files it needs. Architect mode splits the work: one model proposes the change and an editor model writes the file edits. | Yes Plan mode drafts a strategy; Act mode makes the edits, with per-step approval or auto-approve. |
| Agent runs terminal commands | Yes The model suggests shell commands (install dependencies, run tests, run the program) and aider runs them after you confirm; on by default, controlled by --suggest-shell-commands. /run, /test and --auto-test also run commands and feed output back. | Yes |
| Cloud / background agents | Not documented Not documented on vendor pages | Yes Runs on your own machines, not in vendor cloud: CLI Zen mode, cron-scheduled agents and parallel Kanban agents run through a local hub process, and headless runs work in CI. Not available in the VS Code or JetBrains extension. No vendor-hosted cloud agents are documented. |
| Pull request code review | Yes The documented workflow is manual, in the chat: load a branch diff with /run git diff a..b (or --read a diff file), then ask aider about it. Not an automated GitHub PR reviewer. | Yes A GitHub Actions workflow you set up yourself with the Cline CLI, paying only for inference. Not a separate managed review product. |
| Codebase indexing / semantic code search | Yes Repo map: the key classes, functions and signatures of the whole git repo, ranked by a file-dependency graph and trimmed to a token budget (--map-tokens, default 1k), sent with each request. It is a symbol map, not an embeddings search. | No By design, Cline reads files and runs ripgrep-powered search instead of building an index or embeddings. |
| MCP (Model Context Protocol) servers | Not documented Not documented on vendor pages | Yes Includes an MCP Marketplace for one-click installs. Enterprise can allowlist MCP servers. |
| Project rules / custom instruction files | Yes A conventions markdown file (e.g. CONVENTIONS.md), loaded read-only with /read or --read. The read: setting in .aider.conf.yml loads it every session. | Yes Also supports global custom instructions and Skills. |
| Checkpoints to roll back agent edits | Yes Aider git-commits each of its edits; /undo discards the last aider commit. This depends on git integration, which is on by default. | Yes Enabled by default; kept in a shadow Git repository separate from your project history. |
| Aider | Cline | |
|---|---|---|
| User can choose the AI model | Yes Choose with --model at launch or /model during a chat. Separate editor and weak models are configurable too. | Yes |
| Model families offered | Almost any LLM via LiteLLM; documented providers: OpenAI, Anthropic, Gemini, Groq, LM Studio, xAI, Azure, Cohere, DeepSeek, Ollama, OpenRouter, GitHub Copilot, Vertex AI, Amazon Bedrock, OpenAI-compatible APIs The LLMs page recommends Gemini 2.5 Pro, DeepSeek R1 and V3, Claude 3.7 Sonnet, and OpenAI o3, o4-mini and GPT-4.1. | Claude, GPT, Gemini, open-weight models (GLM, Kimi, DeepSeek, MiniMax, MiMo, Qwen), local models, any OpenAI-compatible endpoint The ClinePass page lists open-weight models from Z.ai (GLM), Moonshot AI (Kimi), DeepSeek, MiniMax, MiMo and Qwen. The pricing FAQ adds provider routes: OpenRouter, AWS Bedrock, GCP Vertex, Groq, Cerebras and Vercel AI Gateway. The docs list 30+ further providers. |
| Bring your own API key | Yes Your own provider key is how most models are used; set it on the command line, in environment variables, in a .env file or in .aider.conf.yml. A GitHub Copilot token also works, billed through your Copilot subscription. | Yes Available on the free open-source plan. |
| Local / self-hosted models | Yes Ollama, LM Studio, or any local server with an OpenAI-compatible API. | Yes |
| How paid-plan usage is limited | Not documented Not documented on vendor pages | Pay-as-you-go inference credits at model cost, or your own API keys; ClinePass quota measured over 5-hour, weekly and monthly windows The pricing page lists "Purchase and Model Inference at Cost or BYOK". ClinePass ($9.99/month) measures usage against a 5-hour rolling window, the calendar week and the calendar month. |
| Open-source client | Yes Apache 2.0, source on GitHub (Aider-AI/aider). | Yes Apache-2.0. The GitHub README says the JetBrains plugin is not open-sourced. |
| Aider | Cline | |
|---|---|---|
| Zero data retention option | Yes Default for everyone: aider runs on your machine and the vendor never collects code, prompts, chats or keys. Its analytics are opt-in and exclude them. Code and prompts go to the LLM provider you configure, whose retention terms apply. | Yes Applies when you use your own API keys and route model calls through infrastructure you control, on any plan including the free one. Your model provider's retention terms still apply. No zero-retention commitment is documented for Cline-managed inference. |
| Code excluded from model training | Not documented Not documented on vendor pages | Yes Conditions: Terms §3.5 say third-party AI model providers "may use User Content for training unless you have explicitly opted out". The free-models docs say free-model usage "may be used to help improve model performance and quality". |
| Self-hosted or on-premises deployment | Yes Aider always runs on your own machine or container (pip, uv, pipx or two Docker images). With local models (Ollama, LM Studio, OpenAI-compatible servers), inference also stays in your own infrastructure. | Yes Enterprise. Inference can run through your own Amazon Bedrock, GCP Vertex, Azure OpenAI or local models. |
| Team admin dashboard and usage analytics | Not documented Not documented on vendor pages | Yes Enterprise plan. The enterprise docs list per-team cost breakdowns, real-time analytics and OpenTelemetry export. |
| Aider | Cline | |
|---|---|---|
| Listed integrations | Not documented Not documented on vendor pages | 199 MCP server cards in the "All" tab of the vendor's public MCP Marketplace (198 distinct names). The page states no total. Servers install through the Cline extension, and any other MCP server can be added by hand. |
| Public API | Yes Command line: --message or --message-file runs one instruction and exits. Python: Coder.create() and coder.run(). The vendor says the Python API is not officially supported and may change without backward compatibility. | Yes TypeScript SDK (@cline/sdk). Also an OpenAI-compatible Cline API for model access and an Enterprise admin REST API. |
| Single sign-on (SSO) | Not documented Not documented on vendor pages | Yes Enterprise plan. SSO runs through WorkOS (Okta, Entra ID, Google Workspace, generic SAML/OIDC). The pricing FAQ also lists SCIM provisioning. |
| Two-factor authentication | Not documented Not documented on vendor pages | Not documented Not documented on vendor pages |
| SOC 2 report | Not documented Not documented on vendor pages | Yes Listed under "Compliance" on the vendor Trust Center (Vanta). The Trust Center names no report type or date and offers no downloadable SOC 2 report (resources: Engagement Letter, Code of Conduct). |
| ISO 27001 certification | Not documented Not documented on vendor pages | Not documented Not documented on vendor pages |
| HIPAA compliance offered | Not documented Not documented on vendor pages | Not documented Not documented on vendor pages |
| Aider | Cline | |
|---|---|---|
| Support channels | GitHub issues, Discord, in-chat /help /help answers questions about using aider inside the chat, and /report opens a GitHub issue. Privacy questions go to privacy@aider.chat. | Support contact form, email (support@cline.bot), Discord community, GitHub Issues and Discussions; dedicated support and SLA on Enterprise The pricing page lists "Community Support" for the open-source plan and "SLA" and "Dedicated Support" for Enterprise. The Terms give support@cline.bot. |
| 24/7 support | Not documented Not documented on vendor pages | Not documented Not documented on vendor pages |
| Vendor (legal entity) | Aider AI LLC Named in the FAQ and in the privacy policy. The LICENSE file names no copyright holder (the Apache template placeholder is left as is). | Cline Bot Inc. Terms name Delaware as the governing jurisdiction. |
| Founded | Not documented Not documented on vendor pages | Not documented Not documented on vendor pages |
| Headquarters | Not documented Not documented on vendor pages | Not documented Not documented on vendor pages |
| Aider plan | Billed annually | Billed monthly |
|---|---|---|
| Open source (Apache 2.0) Free; LLM usage billed by your model provider The only way to get Aider: the open-source tool under the Apache 2.0 licence, which grants a no-charge, royalty-free licence. The vendor lists no paid plans (aider.chat/pricing returns 404). Most models need an API key from an LLM provider, and you pay that provider for usage. Free options: OpenRouter free models (daily usage limits) and local models. | Free | Free |
| Cline plan | Billed annually | Billed monthly |
|---|---|---|
| Open Source Free; AI inference billed separately (pay as you go) For individual developers: VS Code extension, CLI, MCP Marketplace, community support. The client is free; model usage is paid per use with Cline credits at cost or with your own provider API keys. | Free | Free |
| Enterprise Custom (contact sales) Price listed as "Custom". Adds JetBrains extension (per the pricing page), SSO, SLA, dedicated support, centralized billing, simple config management, role-based access control, limiting inference providers, team management dashboard and authentication logs. | Custom | Custom |
| ClinePass Per month Optional subscription on its own page (not on the pricing page) for curated open-weight models from Z.ai, Moonshot AI, DeepSeek, MiniMax, MiMo and Qwen, with 2-5x the usage of standard API rates; quota measured over a 5-hour rolling window, the calendar week and the calendar month. Only monthly billing is listed. Page footnotes: "Standard rate $9.99 per month after promotion period" and "Additional processing fee may apply". | Not offered | $9.99 |
Per user per month unless the plan states otherwise, as listed on each vendor's pricing page.