₹649 Cursor Start Tier Targets India Ahead of SpaceX Merger
Cursor's new ₹649 localized subscription for Indian developers includes Grok 4.5 and cloud agents but removes external frontier models and advanced automations.
On July 28, 2026, Anysphere introduced Cursor Start, a localized subscription tier priced at ₹649 per month for developers in India. The plan reduces the barrier to entry in the company’s third-largest global market, where the user base has tripled over the past year. By stripping out select premium tools and external models, the new tier costs approximately $7, significantly less than the standard $20 global Pro tier.
Pricing and Feature Tradeoffs
To achieve the ₹649 price point, Cursor Start restricts model access and advanced automation features. The plan relies exclusively on first-party and affiliated models, explicitly excluding frontier models from OpenAI and Anthropic.
The subscription provides access to Grok 4.5, positioned as the primary powerhouse model for complex reasoning, alongside Composer 2.5 for price-efficient, low-latency coding tasks. To maintain infrastructure margins at the lower price point, Cursor removed compute-heavy UIs and experimental workflows.
| Feature | Cursor Pro ($20) | Cursor Start (₹649) |
|---|---|---|
| Core Models | Claude 3.5, GPT-4o, Grok 4.5 | Grok 4.5, Composer 2.5 |
| Payment Methods | Standard | Cards, Unified Payments Interface (UPI) |
| Auto Mode | Included | Excluded |
| Bugbot | Included | Excluded |
If you rely on the Cursor SDK or build complex continuous Automations, you will need to maintain a Pro subscription, as these specific frameworks are excluded from the Start tier.
Agentic Capabilities
Despite the feature cuts, Cursor Start retains the core autonomous features necessary for modern development workflows. The plan includes always-on cloud agents capable of building, testing, and opening Pull Requests autonomously in the background.
Developers also receive support for Cursor for iOS, which allows remote control and steering of active agents directly from mobile devices. The tier maintains full compatibility with Model Context Protocol (MCP) servers, allowing developers to connect custom tools, plugins, hooks, and skills to their local environment. If you build custom context integrations via MCP, this architecture remains fully functional on the localized plan, ensuring that local repository context and external API data still flow into Grok 4.5.
Strategic Context and Market Positioning
The launch of a localized Indian pricing tier arrives just weeks before the expected close of Cursor’s $60 billion all-stock acquisition by SpaceX. With over 3 million developers in India, the rollout is positioned to lock in market share in the world’s second-largest developer pool before SpaceX absorbs the company.
At ₹649 per month, Cursor Start sits competitively between entry-level consumer AI and premium developer tools. It significantly undercuts Claude Pro, which runs approximately ₹2,000 per month, while offering more developer-specific functionality than the ₹399 monthly ChatGPT Go plan. By supporting UPI payments natively, the platform removes a significant frictionless billing hurdle that often impacts global SaaS adoption in the region.
To support the rollout, Cursor is hosting a local community roadshow, starting in Hyderabad on July 30 and moving to Bangalore on August 1. The company has indicated that this Indian release will serve as the architectural and economic template for localized pricing in other emerging markets.
If you manage engineering teams in India, the Start tier provides a cost-effective way to standardize on Cursor’s agentic workflows, provided your pipelines do not depend on external frontier models or automated Bugbot reviews.
Get Insanely Good at AI
The book for developers who want to understand how AI actually works. LLMs, prompt engineering, RAG, AI agents, and production systems.
Keep Reading
How to Control Token Budgets in Claude Code via Effort Levels
Learn how to configure model selection and the new effort parameter in Claude Code to optimize token spend and agentic task thoroughness.
$230 Codex Micro Macropad Controls OpenAI Coding Agents
OpenAI has released the Codex Micro, a $230 physical macropad developed with Work Louder for managing multi-agent coding workflows.
Grok Build 0.2.93 Caught Uploading Full Git Repositories
An independent wire-level analysis revealed that xAI's command-line tool was silently uploading entire codebases to Google Cloud Storage.
Meta Prices 1M-Token Muse Spark 1.1 at $1.25 Per Million Input
Meta's Superintelligence Labs has launched Muse Spark 1.1, a multimodal reasoning model for agentic workloads, alongside its first metered developer API.
How to Manage Claude Code Workflows via CLAUDE.md
Learn how to control Claude Code's agentic behavior using CLAUDE.md files, reusable skills, lifecycle hooks, and isolated subagents.