What Cursor's $7 India Plan Teaches AI SaaS Founders About Geographic Pricing

Key takeaways

• Cursor launched "Cursor Start" at ₹649/month (about $7), roughly a third of its $20/month Pro plan, exclusively for India.
• India is reportedly Cursor's third-largest market, and its user base there has more than tripled over the past year.
• The cheaper plan excludes frontier third-party models (from providers like OpenAI and Anthropic) and premium features, running instead on Cursor's own Composer 2.5 and Grok 4.5, which the company says keeps it commercially sustainable rather than a loss leader.
• Anthropic and OpenAI have both already localized pricing for India, so this is now a competitive baseline in the market, not a one-off experiment.
• The mechanism, routing a lower-price tier to lower-COGS models, is a repeatable pattern any AI SaaS founder can model before copying.

Why did Cursor launch an India-only $7 plan?

Because India is now Cursor's third-largest market globally, its user base there has more than tripled in a year, and the company's head of APAC and Japan called it "an opportunity to right-size the commercial model and drive scale." That's straightforward geographic price discrimination: charge what a market can bear rather than run one global number for everyone.

What actually makes a much cheaper plan sustainable?

The lever is model cost, not just a marketing discount. Cursor Start runs on Cursor's own Composer 2.5 and Grok 4.5, and excludes access to the frontier third-party models behind its $20 Pro tier, along with premium features like Bugbot, Auto Mode, Automations, and the SDK. Cursor's APAC lead said the plan is viable because it is "built around Cursor's own AI models, which carry lower operating costs than relying primarily on third-party frontier models." Say, for example, your cost on a frontier model runs several times higher per request than your own smaller or fine-tuned model: a tier built on the cheaper backend can absorb a much lower price and still hold margin.

Is localized pricing only for giant companies?

Not necessarily, though it helps to have the scale to justify separate billing infrastructure and payment rails (Cursor built support for India's UPI and rupee billing). OpenAI ran a similar playbook with its sub-$5 ChatGPT Go plan in India before expanding it to 16 more countries, and Anthropic already localizes Claude pricing for India too. The pattern, start with your highest-volume price-sensitive market, route it to your cheapest viable backend, expand once proven, is copyable by smaller teams once you know which market and which model tier can carry it.

What should you check before copying this move?

Model the actual margin at the lower price point using your real cost for the cheaper backend, not your blended average across all tiers. A geographic discount only works if the model behind it is cheap enough to keep the unit economics positive, otherwise you are subsidizing growth in a market you can't sustain.

Geographic price cuts only work when paired with a genuinely cheaper backend, model the unit economics before you localize. You can simulate this tier by tier in Calcaas.