SevenTnewS

Product Launch

Cursor's India plan: model router picks cheapest AI, 649 INR/month

Cursor Start gives Indian developers local pricing with UPI, while Cursor Router lets teams control the cost-intelligence tradeoff per request. Two features that lower the barrier and optimize spending.

Emmanuel Fabrice Omgbwa Yasse AI-assisted

2026-08-01 · 2 min read

Cursor's India plan: model router picks cheapest AI, 649 INR/month

Cursor is making two moves with one release. First, a dedicated pricing tier for India at 649 INR per month (about $7.80), with UPI support. Second, a model router called Cursor Router that decides which model handles each request based on cost and intelligence. Both launch July 28, following a pattern of cost-cutting experiments in AI coding tools, as seen with OpenCode's cheaper Gemini tier that slashes costs by 80%.

The India plan, Cursor Start, is aimed at developers who want agentic coding without the $20/month Pro price. It includes access to Grok 4.5 and Composer (no fast mode), cloud agents that run 24/7, and iOS remote control. But the more interesting piece is the router, which comes with the plan and is also available on Teams and Enterprise. Cursor has been steadily expanding its mobile and chat features, as detailed in parallel chats and mobile agents.

Cursor Router classifies each query by task type and complexity, then sends it to the most appropriate model. Administrators choose from three optimization modes: Intelligence (highest quality, costs more), Balance (high quality, moderate cost), Cost (good quality, cheapest). The router takes a Pareto approach to the cost-intelligence frontier. For the Cost mode, Cursor requires Grok 4.5 as a routing option. This orchestration layer mirrors a broader industry trend in which model routing and switching become more important than raw model size, as discussed in the shift from model size to orchestration.

This is a direct response to a problem that has been nagging agentic coding tools: developers want powerful AI but don't want to burn through credits on simple requests. Cursor's router lets them set a policy once, and the infrastructure handles the tradeoff per request. It is available on desktop, web, iOS, CLI, and SDK. Similar cost discipline has already yielded savings at scale, with Microsoft's in-house AI models cutting production costs by 89%.

The update also brings parallel chats, agent transcript search, and improved project selectors. In Slack, Cursor now shares its plan before executing, runs in multi-repo environments, and can read other channels. But the headline features are the ones that change how developers pay and what they get for the money.

For Indian developers, the local pricing removes a barrier. 649 INR is roughly the cost of a monthly bus pass in most Indian cities. Combined with UPI, it makes the subscription feel native rather than a dollar-denominated upcharge. That is a play for volume in a market where developer tooling is increasingly competitive.

Cursor Router, meanwhile, is a feature that could matter for teams anywhere. It is not just about saving money: it is about maintaining velocity by not forcing every request through the most expensive model. The Balance mode, for example, aims for the quality of the top models most developers use daily without the premium price tag. That kind of fine-grained control has been missing from most AI coding assistants. The same strategic thinking about model orchestration appears in national AI sovereignty plans, such as Japan's orchestration-first approach.

Get the tech essentials in 3 minutes every morning

One email, every weekday, with what actually matters in AI and tech.