Codex and Cursor are not always an either-or choice: OpenAI documents running the Codex IDE extension inside Cursor. Compare the editor and agent paths, execution controls, usage billing, and private-code policy for coordinated repository changes. This guidance is based on company pages linked from the profiles below.
You want one coding agent across local repository work, cloud delegation, terminal, editor, and code review workflows; the CLI can also use local OSS providers through Ollama or LM Studio.
Look elsewhere if: You need managed cloud agents to run on infrastructure you control, need a fixed task allowance, or want cloud tasks while authenticating only with an API key.
You want repository chat, agents, and multi-file editing in Cursor's editor, from its CLI in an existing IDE workflow, or through a documented ACP integration.
Look elsewhere if: You need Cursor inside an editor beyond the documented JetBrains, Zed, or Neovim paths without building a custom ACP client, need confirmed JetBrains ACP access on Cursor's free Hobby plan, or cannot route code context through Cursor's hosted backend.
Free and Go include limited trial access. Plus is $20 per month. Pro starts at $100 per month, with current tiers offering 5× or 20× Plus usage. Business is $20 per user monthly when billed annually or $25 month to month, with at least two users. Enterprise and Edu use sales-led pricing. OpenAI's newer rate card says Plus, Pro, and Business accounts plus most Enterprise-family accounts use token-based rates, while its Codex pricing page still says only Business and new Enterprise accounts have migrated; check the Usage panel for the meter that applies to your account. Some Plus and Pro users can add credits, while others may need to upgrade or wait for a reset. All users can run extra local tasks with an API key at standard API rates. Credit use depends on the model, input and output token mix, speed, and any additional agents. OpenAI publishes a cross-plan planning estimate of roughly $100–$200 per developer per month, with large variance; it does not define an included allowance or predict an individual bill.
Hobby is free. Pro costs $20 monthly and includes $20 of third-party model usage; Pro+ costs $60 and includes $70; Ultra costs $200 and includes $400. On individual plans, third-party model usage is charged at the model's API price. On Teams and Enterprise, Cursor adds $0.25 per million tokens to third-party model API pricing for included, on-demand, and BYOK usage; Auto Cost and Cursor's first-party models are exempt. Cursor describes each plan's separate allowance for its own models only as ‘generous included usage,’ without a number. Teams Standard is $40 per user monthly; Premium is $120 per user monthly and provides 5× Standard’s limits on Agent. Pro, Pro+, and Ultra can continue after included usage with monthly on-demand billing; Start does not include on-demand usage. On Teams, usage is allocated per user and does not transfer; on-demand is enabled by default, with a configurable monthly team-wide spending limit. Cloud Agents are charged at the selected model's API price, and Cursor asks you to set a spend limit before first use.
Ways to use it
Web, macOS, Windows, Linux, iOS, Terminal, VS Code
Available through Codex in ChatGPT, the desktop app, CLI, IDE extension, web, and iOS, subject to plan, platform, and workspace settings. The Codex IDE extension also supports Cursor and Windsurf; OpenAI documents separate integrations for Xcode and JetBrains IDEs. Local CLI, IDE, and SDK use can instead authenticate with an API key; API-key access does not include cloud features such as GitHub code review or Slack.
Desktop editor and terminal, ACP support for JetBrains IDEs, Zed, Neovim, and custom clients, plus cloud agents through a browser, phone, Slack, GitHub, or Linear. The native iPhone and iPad app supports agent monitoring, an inbox, and full pull-request review; Cursor says iPad access is available on paid plans. Cursor says JetBrains ACP is free for users on paid plans; the checked pages do not state whether Hobby users can use it.
Where it runs
Local workflows run commands on the user's device under configurable sandbox, approval, and network policies; the CLI can also use OSS models served by Ollama or LM Studio. Cloud tasks run in isolated OpenAI-managed environments; setup scripts have internet access, while agent internet access is off by default unless configured.
Do you need to choose one, or can Codex run inside Cursor?
OpenAI documents installing the Codex IDE extension in Cursor, so Cursor can be the editor while Codex handles an agent session. OpenAI's page covers extension support, editor context, local work, and optional cloud handoff; compare plan and agent details through the separate product pages below. Choose Cursor's native workspace when its editor, CLI, cloud agents, or documented ACP integrations are the main workflow. Choose Codex's desktop app, CLI, web, iOS, or cloud delegation when those surfaces matter; or test the Codex-in-Cursor path when you want Cursor's editor around Codex.
Do you need a documented local-model or hosted-request path?
Codex CLI documents an OSS-model path through Ollama or LM Studio. OpenAI documents that local-provider route for the CLI. Cursor says requests still pass through its backend for final prompt building, including when you supply your own API key. For a hard local-inference requirement, Codex has a documented CLI path while Cursor's documented request path remains hosted. Verify the exact client and model rather than equating a local editor with local inference.
Which approval, sandbox, and cloud-network boundary fits the repository?
For local work, Codex separates a platform-enforced sandbox from approval policy: routine commands can run inside configured filesystem and network boundaries, while crossing them can require approval. Cursor's desktop Agent can edit workspace files without approval. For Shell, MCP, and Fetch calls, its Auto-review mode runs allowlisted calls immediately, sandboxes calls it can, and uses a classifier to allow the rest, try another approach, or ask for approval; Run Everything removes prompts and sandboxing. Cloud behavior differs too. Codex cloud tasks run in isolated environments where setup scripts have internet access and agent internet access is off by default. Cursor Cloud Agents auto-run terminal commands in isolated virtual machines with internet access on by default. Choose and test the exact client and mode rather than treating either product name as one permission model.
Which produces a mergeable change with less supervision on your repository?
Do not answer this from one benchmark or demo. First decide whether you are comparing the products as sold—using each product's normal default model and mode—or the agent harnesses with the same model version where both offer it. From the same clean commit, give each a bounded task, identical repository instructions and acceptance checks, and the closest comparable permissions, time, and spend ceilings each permits; record every mismatch. Run at least three fresh attempts per setup. That can reveal instability, not establish a universal advantage. For every run, record the date, client version, model, mode, elapsed time, reported usage, and every human intervention; mark missing or non-comparable figures unknown. Then inspect the diff and judge requirements met, tests passed, regressions, and unnecessary edits. The result applies only to that repository, task, setup, and date.
Use the current price row above for subscription costs. ChatGPT Plus, which includes Codex, and Cursor Pro each list a $20 monthly individual plan, but neither promises a fixed repository-task allowance. OpenAI's newer rate card says Plus, Pro, and Business accounts plus most Enterprise-family accounts use token-based Codex rates, while its Codex pricing page still says only Business and new Enterprise accounts have migrated; check the account's Usage panel. Codex also shares one agentic usage and credit pool with other eligible ChatGPT agent features. Some Plus and Pro users can add Codex credits. Cursor's Pro plan lists $20 of Other Models usage priced at those models' API rates and lists Cursor Models as ‘Generous included usage’; Pro can continue with monthly on-demand billing. A comparable fixed repository-task allowance is unknown because the documented usage units differ. For a hard $20 ceiling, leave Codex credit auto-reload off, do not add credits or API billing, and disable Cursor on-demand usage. Then test a representative week and record the client, model, task size, and reported usage instead of converting either plan into an unsupported task count.
How may prompts and private code context be used for model training?
For Plus and Pro, OpenAI may use Codex content to improve models. ChatGPT conversations and Codex tasks can be opted out through general data controls, while full-environment Codex training has a separate control in Codex Settings that the general controls do not change. Business, Enterprise, Edu, and API inputs and outputs are not used for training by default; eligible API organizations can opt in. With Privacy Mode on, Cursor says it and its model providers do not train on customer data; with it off, Cursor may use codebase data, prompts, editor actions, and code snippets for training. Set the relevant account or workspace control before opening private code, and review retention separately rather than treating a training opt-out as a deletion policy.
