Cursor Start India: ₹649 Price, Models and Limits

Cursor StartAn India-only individual plan centered on Cursor Models, with cloud and mobile access but without Other Models, Auto, Bugbot, Automations or the Agent SDK. Press Enter for the full definition. is a monthly individual plan for developers in India. It launched on July 28, 2026, between Free and Pro, with local INR billing and support for UPI, credit card, or debit card. The listed ₹649 price includes tax, while Cursor says prices for its other individual plans are shown before tax. That makes the useful comparison the amount charged, not a currency conversion against the Pro headline.

The plan is tied to eligibility in India. Existing Free users can upgrade from the Cursor dashboard, and new users can choose Start while signing up. Cursor has not described it as a discounted Pro plan. It is a different bundle built around the Cursor ModelsCursor's plan and usage pool for models it labels as Cursor Models, kept separate from the third-party Other Models pool. Press Enter for the full definition. pool, so check the included products before treating the lower price as the only difference.

Start includes generous usage from the Cursor ModelsCursor's plan and usage pool for models it labels as Cursor Models, kept separate from the third-party Other Models pool. Press Enter for the full definition. pool. The current pool contains Grok 4.6A Cursor and SpaceXAI frontier model for complex coding and knowledge work; it is a selectable model and is separate from the Grok Bot product. Press Enter for the full definition., Grok 4.5, and Composer 2.5The current Composer release, better at long-running tasks and at judging when a job needs a light touch versus deep work. Press Enter for the full definition.. All three run in Standard mode on this plan. Grok 4.6 and Grok 4.5 are fixed at medium effort, so Start users cannot select low, high, or xhigh effort and cannot turn on Fast mode.

That is enough range for ordinary coding work. Composer 2.5The current Composer release, better at long-running tasks and at judging when a job needs a light touch versus deep work. Press Enter for the full definition. is the lower-cost execution model, while the two Grok versions are available when a task needs more depth. The restriction matters when your workflow depends on a specific speed or reasoning control. In that case the problem is plan capability, not a hidden editor setting, and the documented route is Pro or higher.

Cursor does not publish a numeric Start token allowance. Its documentation calls the allowance generous and says it is intended for daily building. That wording cannot be converted into a dependable request count because tasks consume different amounts of context and output. Watch the usage dashboard with your own workload rather than planning around a number from an older plan or an informal forum post.

Start includes cloud agents and Cursor for iOS, along with the extension layer that lets an agent work with more than repository files. Plugins, MCPModel Context Protocol. A standard that lets an AI agent pull in context from outside the repo, like Jira tickets or internal docs. Press Enter for the full definition. servers, hooks, and skills are all included. In practice, that means the plan can support a local coding session, a long cloud run, or an agent started from a phone without moving to Pro first.

Those products still keep their own setup and security boundaries. A cloud agent needs a connected repository and an environment it can run. An MCPModel Context Protocol. A standard that lets an AI agent pull in context from outside the repo, like Jira tickets or internal docs. Press Enter for the full definition. server gets only the accounts and permissions you authorize. Hooks can inspect or block actions only when configured to do so. The plan grants access to each surface; it does not preconfigure a safe workflow for your project.

1. 1Upgrade or choose Start with an eligible India account.
2. 2Open the model picker and confirm Grok 4.6A Cursor and SpaceXAI frontier model for complex coding and knowledge work; it is a selectable model and is separate from the Grok Bot product. Press Enter for the full definition., Grok 4.5, and Composer 2.5The current Composer release, better at long-running tasks and at judging when a job needs a light touch versus deep work. Press Enter for the full definition. appear in Standard mode.
3. 3Connect one test repository before starting a cloud agent.
4. 4Add plugins or MCPModel Context Protocol. A standard that lets an AI agent pull in context from outside the repo, like Jira tickets or internal docs. Press Enter for the full definition. servers one at a time and review their account permissions.
5. 5Check the usage dashboard after a representative week of work.

A small trial is useful here because included access and useful access are different questions. If your normal day stays inside ComposerCursor's own fast coding model, tuned for the editor and priced well below frontier models; the recommended day-to-day model for executing a plan. Press Enter for the full definition. and an occasional Grok run, Start may fit well. If every task needs a third-party model or an automation trigger, the missing product will show up immediately and you can compare Pro without guessing.

Start does not include the Other ModelsCursor's pool of supported third-party models from providers such as Anthropic, Google and OpenAI, with separate plan access and usage from Cursor Models. Press Enter for the full definition. pool, on-demand usage, BugbotCursor's automated PR reviewer that posts inline findings and can push fix commits from isolated VMs. Press Enter for the full definition., Auto, Automations, or the Cursor SDK. The first limit removes directly selected models from Anthropic, Google, OpenAI, and other outside providers. The remaining limits remove automated model routing, pull request review by Bugbot, scheduled or event-driven agent runs, and programmatic agents built with the SDK.

Paid overflow is the less visible difference. Pro can continue past included limits through on-demand usage when enabled; Start cannot. If a deadline depends on uninterrupted capacity, a generous allowance is still a hard operational boundary once it runs out. Check your dashboard before a long batch rather than assuming the plan will bill the extra work automatically.

Move to Pro when a missing capability is part of the job, rather than because another plan has a longer feature list. Clear triggers include needing Claude, Gemini, or GPT models, choosing Fast or a different Grok effort level, running BugbotCursor's automated PR reviewer that posts inline findings and can push fix commits from isolated VMs. Press Enter for the full definition., creating Automations, building with the Cursor SDK, or continuing with paid usage after an included pool is exhausted.

Before moving, save one month of usage history and list the requests that Start could not complete. That record separates a real plan limit from a prompt, environment, or repository problem that would follow you to Pro. It also gives you a concrete way to decide later whether the higher tier is still earning its price.

If none of those conditions applies, compare actual monthly use before upgrading. Start already covers desktop, web, CLI, mobile, and cloud agent work with Cursor's current model pool. The decision is fairly plain: pay for Pro when the excluded tool changes an outcome you care about, and stay on Start when the included workflow produces the same checked result.