Cursor Dev Stack
BeginnerCursor as editor and coding agent in one, with Grok by default and Claude, GPT, or Gemini a click away.
Published 27 September 2026
About Cursor Dev Stack
Cursor is an AI-native code editor built on VS Code, and it is also a coding agent: the same app that gives you Tab completions and a file tree can take a plain-language task, plan it, edit files across the codebase, and run terminal commands to check the result. On paid plans, cloud agents run the same agent in Cursor's own cloud machines, so a task keeps going with your computer off; you start and follow them from the web, the desktop app, the iOS app, Android through the web app, Slack, or a comment on GitHub or Linear. For most developers the editor and the agent are one product, which is why this stack has a single Cursor at its core rather than an editor plus a separate agent.
Models are picked per chat from Cursor's own model picker. Grok is the default here, with generous Grok access included in Cursor's paid plan; Claude, OpenAI's GPT models, and Gemini sit in the same picker, and Cursor's own Composer model is included on every plan. Several can be in use at once: a fast model for everyday edits, and a stronger one for a hard refactor in another chat.
A separate coding agent is an optional extra, not a replacement. Because Cursor runs VS Code extensions, you can optionally install the official Claude Code or Codex extension in its sidebar, or run an agent CLI such as OpenCode in Cursor's integrated terminal if you want one. That keeps Cursor's editor and Tab completions while another vendor's agent takes on a task. For version control, GitHub is the default, and Cursor pulls its pull-request context straight into chat; GitLab fits when source control has to be self-hosted.
Docker and a CI/CD pipeline are common next additions once a project outgrows one laptop: Docker for environment parity between development and production, and GitHub Actions or GitLab CI/CD for tests that run on every push. Neither is needed to get started.
Key Features
- ✓One app for editing and agentic work: Tab completions, chat, and an agent that edits files and runs commands
- ✓Grok by default, with Claude, GPT, and Gemini in the same model picker, chosen per chat
- ✓Cloud agents that keep working on background tasks after the laptop is closed
- ✓Optional Claude Code and Codex extensions, or any agent CLI in the integrated terminal
- ✓VS Code extension and keybinding compatibility, so existing setups carry over
- ✓Pull-request context from GitHub pulled straight into chat
When to Use Cursor Dev Stack
- →Scaffolding a feature by describing it and letting the agent edit every file it touches
- →Switching between a fast model for small edits and a stronger one for a hard refactor
- →Handing a long-running task to a cloud agent while you keep working in the editor
- →Trying Claude Code or Codex on a task without leaving the editor you already use
- →Moving from VS Code to an AI-native editor without losing extensions or keybindings
Pros
- Editor and agent in one product, with no separate tool to install to get started
- Model choice across Grok, Claude, GPT, and Gemini from one subscription
- Built on VS Code, so existing extensions and keybindings keep working
- Other vendors' agents fit in through extensions or the terminal when you want them
Cons
- Heavy use of premium models runs through plan limits faster than the default model
- A separate agent like Claude Code or Codex needs its own subscription on top of Cursor's
- Agent-written code still needs review; confident but wrong edits can slip through
- Closed-source editor, so the workflow depends on Cursor's product decisions
LLM Options for Cursor Dev Stack
Cursor Dev Stack powered by Grok
Grok is the default model here: xAI's models are Cursor's headline option, with generous Grok usage included in the paid plan, and they're fast enough for everyday edits and agent tasks. Cursor's own Composer model, included on every plan, sits alongside it in the picker.
Cursor Dev Stack powered by Claude
Claude models are a common pick for large multi-file changes and careful reasoning about unfamiliar code. Switch to Claude per chat when a task needs more deliberation than speed; premium models draw down Cursor's plan usage faster than the default, so heavy use fits the Pro+ or Ultra tiers.
Cursor Dev Stack powered by OpenAI
OpenAI's GPT models sit in the same picker, useful for a second opinion when one model gets stuck, or for teams already standardized on OpenAI. They're billed through Cursor's plan, so there's no separate OpenAI account to manage.
Cursor Dev Stack powered by Google Gemini
Gemini's long context window suits questions that span the whole repository, like tracing a data flow or planning a large migration, and its Flash models are a fast, lower-cost pick for routine edits inside Cursor's plan limits.
These are highlighted picks. To see all the tools, check the LLM category.
Version Control Options for Cursor Dev Stack
Cursor Dev Stack with GitHub
GitHub is the default here: the largest pull-request and Actions ecosystem, and Cursor's GitHub integration pulls pull-request context straight into chat and the agent.
Cursor Dev Stack with GitLab
GitLab bundles source control, merge requests, and CI/CD in one self-hostable platform, the natural pick for teams standardized on GitLab or that need code and pipeline data to stay on their own infrastructure.
Cursor Dev Stack Add-ons
Each addition below extends this stack with a capability the base stack works fine without. None are required: include the ones your product actually needs when building this stack, and skip the rest.
Coding Agent Add-ons
Add a separate coding agent when you want another vendor's agent working alongside the one built into the editor, either through its own extension or in the integrated terminal.
Cursor Dev Stack with Claude Code
Add Claude Code when you want Anthropic's agent on a task without leaving Cursor: its official extension installs in the sidebar, or run the CLI in the integrated terminal. It needs its own Claude plan.
Cursor Dev Stack with Codex
Add Codex when you want OpenAI's agent alongside Cursor's: the official Codex extension works in Cursor as it does in VS Code, and it runs on a ChatGPT plan you may already have.
Cursor Dev Stack with OpenCode
Add OpenCode in Cursor's integrated terminal when you want an open-source agent that runs any provider's model with your own API key, including cheaper open-weight ones.
These are highlighted picks. To see all the tools, check the AI Coding Agents category.
Code Review Add-ons
Add code review when you want an AI reading every pull request before it merges: it comments on the diff so bugs, security issues, and inconsistencies surface before a human has to catch them.
Cursor Dev Stack with Cursor Bugbot
Add Cursor Bugbot to keep review inside the Cursor plan this stack already pays for: it reviews GitHub or GitLab pull requests for bugs specifically, billed from the plan's usage at about $1 to $1.50 per review.
Cursor Dev Stack with CodeRabbit
Add CodeRabbit when you want a broader reviewer than Bugbot alongside the editor: it walks through the whole pull request on GitHub or GitLab with a summary and line-level comments, with a free tier for open-source repos. Its IDE extension also runs inside Cursor.
Cursor Dev Stack with Greptile
Add Greptile when you want review that reads the whole repository rather than just the diff: it indexes the codebase so its GitHub and GitLab pull request comments carry wider context, and it can be self-hosted.
CI/CD Add-ons
Add CI/CD when you want a dedicated pipeline for running tests, linting, or multi-stage builds before a deploy goes out. Many hosting platforms already redeploy automatically on every push on their own — a CI/CD tool adds the most value on top of that by gating the deploy on a passing test suite, and matters even more when the hosting choice does not auto-deploy at all, such as a self-hosted server.
Cursor Dev Stack with GitHub Actions
Add GitHub Actions when the repo lives on GitHub and you want tests, linting, and deploy jobs running automatically on every push.
Cursor Dev Stack with GitLab CI/CD
Add GitLab CI/CD when the project is on GitLab and needs the same automated pipeline, defined in a .gitlab-ci.yml file that can run entirely on your own infrastructure.
These are highlighted picks. To see all the tools, check the CI/CD Pipelines category.
Containerization Add-ons
Add containerization when you want the app packaged the same way across local development, staging, and production, or need to deploy somewhere that isn't a managed serverless platform.
Cursor Dev Stack with Docker
Add Docker when the project needs environment parity between a developer's laptop and production, or when onboarding a new teammate should be a single docker compose up instead of a page of setup instructions. Cursor's agent can write the Dockerfile and Compose config when asked.
These are highlighted picks. To see all the tools, check the Containerization category.
Frequently Asked Questions about Cursor Dev Stack
Can Cursor keep working when my computer is off?
Yes, through cloud agents on a paid plan. They run in Cursor's own cloud machines with the repository cloned, so your computer can be off, and they hand back a branch or pull request. You can start and check on them from cursor.com, the desktop app, the iOS app, Android through the installable web app, Slack, or an @cursor comment on GitHub or Linear. They work with GitHub and GitLab repositories alike. Cloud agents are billed at API pricing for the model you pick, on top of the plan, so long runs on premium models add up.
Why is Grok the default model?
Grok is Cursor's headline model option, with generous Grok usage included in the paid plan, and Cursor's parent company is being acquired by SpaceX, which merged with xAI. It's only a default, though: the model picker switches to Claude, GPT, or Gemini per chat, and Cursor's own Composer model is included on every plan, free tier included.
Can I use Claude Code or Codex inside Cursor?
Yes. Cursor runs VS Code extensions, so the official Claude Code and Codex extensions install into its sidebar, and any agent CLI can run in the integrated terminal. Each needs its own subscription, since Cursor's plan only covers its own agent and models.
Can I use several models in the same project?
Yes. The model is chosen per chat or agent task, so one chat can run Grok for quick edits while another uses Claude for a hard refactor. Auto mode picks a model for you. Premium models draw down plan usage faster than the default, which is the practical limit on mixing them.
How is this different from stock VS Code with an AI extension?
Cursor is a fork of VS Code with the agent, Tab completions, and model picker built into the editor itself. Stock VS Code gets AI through an extension such as Copilot or Cline, which can be swapped freely but works within what an extension is allowed to do. Pick Cursor for the most integrated experience, or VS Code to keep the editor unchanged and choose the AI separately.
Scores
Popularity5/5
Cursor is the most widely adopted AI-native editor, known to developers well outside AI-focused teams.
Learning Curve1/5
Anyone who has used VS Code is productive on day one: the agent and chat work from plain-language requests, and adding a separate agent later is an extension install.
Flexibility4/5
Four model providers in one picker, optional third-party agents through extensions or the terminal, and full VS Code extension support, within a closed-source editor whose core behavior is Cursor's to decide.
Performance4/5
Tab completions are near-instant and cloud agents move long tasks off the local machine, though large-context requests on big codebases slow responses down.
Portability4/5
Projects are plain files and git history, and the editor keeps VS Code's settings and extensions, so moving to another editor or agent costs little beyond Cursor-specific rules and habits.
Tools in the Cursor Dev Stack Stack
Cursor Dev Stack Pricing
Cursor's own subscription is the main cost: the free Hobby tier covers light use and includes Cursor's Composer model, the Individual plan at $20/month extends agent limits and adds Grok, frontier models, and cloud agents, and Teams is $40 per user per month. Cloud agents bill at API pricing for the chosen model. A separate agent like Claude Code or Codex, if added, needs its own plan. GitHub and GitLab are free for individual developers, and Docker and CI/CD have free tiers for typical personal use.
Hobby is free with Composer included; Individual is $20/mo with Grok and frontier models, or $60/$200 for higher usage; Teams from $40/user/mo.
Free tier covers individual devs and small teams; GitHub Team is $4/mo, GitLab Premium is $29/mo for more seats.
Claude Code starts at $20/mo (no free tier); Codex has a free tier, with Plus at $20/mo; both go to $200/mo. OpenCode is free and bills per token.
Docker Desktop is free for personal use; Docker Pro adds team features from $5/mo.
GitHub Actions and GitLab CI/CD both include free monthly minutes before paid tiers kick in.