Indian developers should start with GitHub Copilot Free or Cursor Hobby. If you code daily, GitHub Copilot Pro at $10/month is the best value and includes $15 of monthly credits. Cursor’s India-only Start plan at ₹649/month avoids forex but includes $0 of other-model budget. Claude Pro at ₹1,999/month suits people who live inside Claude Code.
The AI coding market repriced itself in 2026, structurally rather than cosmetically. Request counts are gone. Cursor sells dollar-denominated allowances of other-model usage, GitHub Copilot sells monthly credits, and Anthropic sells rolling usage windows. That makes cross-tool comparison harder, and makes the headline price a floor rather than a ceiling on all three.
Best choices at a glance
| Tool | Best for | India price | Key limit |
|---|---|---|---|
| GitHub Copilot Free | Trying AI coding at zero cost | $0 [USD] | 2,000 completions/mo; Haiku 4.5 and GPT-5 mini only; Copilot CLI |
| GitHub Copilot Pro | Daily coding, best all-round value | $10/user/mo [USD] | Completion cap removed; $15 monthly credits; annual billing not published |
| GitHub Copilot Pro+ | Premium models including Opus | $39/user/mo [USD] | $70 credits; 4x+ Pro usage; audit logs |
| GitHub Copilot Max | Heaviest agent usage | $100/user/mo [USD] | $200 credits; 2.9x+ Pro+ usage; priority new-model access |
| Cursor Start | Indian developers avoiding forex | ₹649/mo tax inclusive, India only [web] | $0 of other-model usage — Cursor’s own models only |
| Cursor Pro | Mixed-model agent work | $20/mo · $16/mo annual, 20% off [USD] | $20 other-model allowance; own-model limits not published |
| Cursor Pro+ / Ultra | Heavy agent workloads | $60/mo · $200/mo, $160/mo annual [USD] | $70 and $400 allowances; agent limits published only as multipliers |
| Claude Pro | Claude Code as the main workflow | ₹1,999/mo · ₹19,900/yr [IN-store] · $20/mo, $200/yr [web USD] | Rolling 5-hour session window plus weekly limits; no fixed message count |
| Claude Max | Long agent sessions, higher output | ₹12,900/mo (5x) · ₹24,900/mo (20x) [IN-store] · from $100/mo | Monthly-only; no annual option |
Last verified: 9 August 2026
Who each tool is for
GitHub Copilot suits developers already inside GitHub and VS Code who want autocomplete plus an agent without changing editors. Its four-tier ladder — Free, Pro, Pro+, Max — scales spend to usage without an enterprise plan.
Cursor suits developers who want the editor itself to be agent-first, and who accept an allowance model where the headline price is a starting point. In India it has one property no competitor matches: a rupee-priced plan.
Claude Pro suits developers whose work is as much reasoning and refactoring as completion — reading a large codebase, planning a migration, writing tests. Claude Code is included in Pro, which is the point.
Pricing in India, including the forex point
This is the most consequential difference between these tools for an Indian buyer.
GitHub Copilot prices in USD and labels it explicitly on its pricing page. Cursor prices in USD for every tier except one. Claude’s web pricing is USD; whether the claude.ai web checkout bills INR from India is unverified, though India is on Anthropic’s supported-countries list.
So for Copilot and most Cursor tiers your bank adds a card forex markup, commonly 2 to 3.5 per cent. Copilot Pro at $10/mo is $10 plus markup, every month, at that day’s rate, a small but real premium that makes budgeting less predictable than a rupee price.
The exception is Cursor Start at ₹649/month, tax inclusive, India only. It appears in Cursor’s documentation rather than on the public pricing page, so you may not find it by browsing. It is a genuine India-specific plan and the only rupee-native option among these vendors’ individual tiers.
Claude Pro at ₹1,999/mo or ₹19,900/yr through the Apple App Store India storefront [IN-store] is another rupee route, GST-inclusive, though web checkout may differ. Claude Max is ₹12,900/mo for the 5x tier and ₹24,900/mo for 20x [IN-store], and Max is monthly-only. For a wider view, see our best AI subscriptions in India roundup.
Strengths and limitations, tool by tool
GitHub Copilot
Strengths: the four-tier ladder is the clearest pricing in this category. Free gives 2,000 completions a month with Haiku 4.5, GPT-5 mini and the Copilot CLI. Pro at $10/user/mo removes the completion cap and adds next edit suggestions, the cloud agent, code review, third-party agents including Claude Code and Codex, model selection, and $15 of monthly credits. Pro+ at $39 adds premium models including Opus, audit logs and 4x or more of Pro’s usage, with $70 in credits. Max at $100 adds priority new-model access, 2.9x or more of Pro+ usage, and $200 in credits.
One detail is worth noticing: the credit allowance exceeds the subscription price at every paid tier. $10 buys $15 of credits, $39 buys $70, $100 buys $200. That is unusually favourable and the strongest argument for Copilot on pure value.
Limitations: USD billing with forex on top. Annual billing availability is not published by the vendor in a verifiable form, so assume monthly. The Free tier’s model list is narrow, Haiku 4.5 and GPT-5 mini only, so Free is a fair test of the workflow and a poor test of the intelligence. Business at $19/seat and Enterprise at $39/seat add nothing for a solo developer.
Cursor
Strengths: Start at ₹649/mo tax inclusive is a real answer to Indian pricing complaints, and being tax inclusive it is the number that leaves your account. Pro at $20/mo, or $16/mo annual for 20 per cent off, is competitive. Pro+ at $60 and Ultra at $200 ($160 annual) scale to heavy agent use, with agent limits given as 3x Pro for Pro+ and 20x Pro for Ultra. Teams Standard is $40/user/mo, Teams Premium $120/user/mo or $96 annual.
Limitations, and these are substantial. Cursor Start includes $0 of other-model usage. The allowance model is dollar-denominated: Pro carries $20, Pro+ $70, Ultra $400, Start zero. On Start you are restricted to Cursor’s own models. If your reason for wanting Cursor is routing work to other frontier models inside the editor, Start does not do that, and the rupee price buys a different product from the one you may have read about.
Cursor’s own-model limits are described as “generous included usage” and are deliberately not quantified by the vendor. Agent limits appear only as multipliers, not absolute figures. And overage falls to usage-based billing, so a heavy month can cost meaningfully more than ₹649 or $20. Set a spend limit before you need one.
Claude Pro and Max
Strengths: Pro includes Claude Code, plus Cowork, Design, Science, Research and Microsoft 365 integration, and Anthropic describes it as at least 5x the usage per session versus free. At $200/yr that is about 15 per cent off the monthly rate, with a rupee route at ₹19,900/yr through the App Store India storefront. Max adds higher output limits and priority access.
Limitations: usage is a rolling five-hour session window plus weekly limits on paid plans, with no fixed message count published. That makes it hard to know in advance whether Pro covers your workload, since you find out by hitting the limit mid-task. Max is monthly-only, so there is no annual saving at the top tier, and at ₹12,900 or ₹24,900 a month that is a lot of unhedged spend. If you are choosing between assistants rather than coding tools, our ChatGPT vs Claude comparison for India covers the broader trade-off.
How to decide
- If you have never used an AI coding tool → GitHub Copilot Free. 2,000 completions a month is enough to form an opinion.
- If you are a verified student → GitHub Copilot Student, free. It is Copilot Student, not full Pro, and eligibility is re-evaluated monthly. Verified teachers and popular open-source maintainers get free Copilot Pro instead.
- If you code most days and want the best value → GitHub Copilot Pro at $10/mo. The $15 credit allowance exceeds the price.
- If forex is your main objection and Cursor’s own models suffice → Cursor Start at ₹649/mo, understanding the $0 other-model budget first.
- If you want premium third-party models inside Cursor → Cursor Pro at $20/mo minimum.
- If you need Opus-class models in your editor → GitHub Copilot Pro+ at $39/mo, cheaper than Cursor Ultra and with $70 in credits.
- If your work is reasoning-heavy and Claude Code is your main interface → Claude Pro.
- If your monthly spend must be predictable → prefer Copilot’s credit model over Cursor’s usage-based overage.
- If you are a two-to-five person team → compare Cursor Teams Standard at $40/user/mo against Copilot Business at $19/seat.
Start free, then upgrade
GitHub Copilot Free at $0 gives 2,000 completions per month, Haiku 4.5, GPT-5 mini and the Copilot CLI. For a student or weekend-project developer writing a few hundred lines a week, that cap is not the binding constraint. The narrow model list is.
Cursor Hobby is free and lets you evaluate the agent-first editor, genuinely different from a plugin in VS Code. Claude Free at ₹0 lets you test Claude’s reasoning on your codebase before deciding whether Pro’s five-hour window fits your rhythm.
Upgrade triggers: leave Copilot Free when model quality, not completion count, limits you; leave Cursor Hobby when you want the agent to run longer tasks; leave Claude Free when you hit session limits more than once a week.
Before paying anything, run the same real task, a refactor you actually need rather than a toy problem, through all three free tiers in one afternoon. The differences show up on real codebases and are invisible on tutorials.
Legitimate ways to pay less
Every route here is officially documented:
- GitHub Copilot Student — free for verified students. Copilot Pro is free for verified teachers and maintainers of popular open-source projects. Eligibility is re-evaluated monthly.
- Cursor Start — ₹649/mo tax inclusive, India-only, documented in Cursor’s docs rather than on the public pricing page.
- Annual billing — Cursor Pro at $16/mo saves 20 per cent, Ultra at $160/mo, Teams Premium at $96/user/mo. Claude Pro at $200/yr saves about 15 per cent.
- Set a usage-based billing cap on Cursor before your first heavy month, since overage is billed on top of the plan.
What does not exist: no documented India discount on GitHub Copilot, no student rate on Cursor in the documented pricing, and no annual option on Claude Max. If you see any of those advertised, check the vendor page directly.
PremiumKing lists Cursor and GitHub Copilot among the items in its AI tools category; Claude is not sold there. As with any marketplace listing, confirm the exact plan, tier and term supplied against the vendor’s own plan pages before buying.
FAQ
Is Cursor’s ₹649 Start plan the same as Cursor Pro in rupees? No. Start is India-only at ₹649/mo tax inclusive, and it includes $0 of other-model usage, meaning Cursor’s own models only. Pro at $20/mo includes a $20 allowance of other-model usage. They are different products at different capability levels.
Which is better value, GitHub Copilot Pro or Cursor Pro? On published numbers, Copilot Pro at $10/mo with $15 of monthly credits wins, since the credit allowance exceeds the price. Cursor Pro at $20/mo ($16 annual) includes a $20 other-model allowance and an agent-first editor that some developers prefer enough to pay double for. Both bill in USD from India.
Does GitHub Copilot Free really have a 2,000-completion limit? Yes. The Free tier is capped at 2,000 completions per month with Haiku 4.5, GPT-5 mini and the Copilot CLI. Pro removes the completion cap and opens model selection, the cloud agent and code review.
Does Claude Pro include Claude Code? Yes. Pro includes Claude Code, along with Cowork, Design, Science, Research and Microsoft 365 integration, and Anthropic describes it as at least 5x the usage per session compared with free. Usage is governed by a rolling five-hour window plus weekly limits rather than a message count.
Can I pay for these tools in rupees from India? Partly. Cursor Start at ₹649/mo is rupee-priced and tax inclusive. Claude Pro and Max have INR prices through the Apple App Store India storefront. GitHub Copilot prices in USD and labels it on the page, as do most Cursor tiers, so a forex markup of roughly 2 to 3.5 per cent applies there.
Will my bill stay at the plan price? Not necessarily with Cursor. Overage beyond the included allowance falls to usage-based billing, so the plan price is a floor rather than a ceiling. GitHub Copilot’s credits are a fixed monthly allowance per tier, and Claude’s limits throttle rather than bill you. If predictable spend matters, that difference should decide your choice.
Last verified: 9 August 2026
Primary sources
- cursor.com/pricing and cursor.com/docs/account/pricing
- github.com/features/copilot/plans and docs.github.com/en/copilot/get-started/plans
- claude.com/pricing and support.claude.com/en/articles/8325606
- anthropic.com/supported-countries