GPU cloud providers in India

India-incorporated cloud infrastructure provider (E2E Networks Limited, NSE: E2E) registered in New Delhi, offering GPU instances from B200 down to L4, primarily targeting Indian startups, AI/ML teams, and enterprises.

Global GPU cloud marketplace with community and secure cloud tiers. Confirmed 16-model catalog via their public pricing page's schema.org structured data (B300, H200, B200, RTX PRO 6000, H100 [NVL/PCIe/SXM], A100 [PCIe/SXM], L40S, RTX 6000 Ada, A40, L40, A6000, RTX 5090, L4, RTX 3090, RTX 4090, A5000). Not India-incorporated — no India region, billing in USD, no Indian GST invoicing.

India-based GPU cloud provider ('Cloud built for AI teams') targeting ML researchers and startups, with per-minute billing, template boot times around 1.8 seconds, and VM boot under 90 seconds. Confirmed catalog: H200, H100, RTX Pro 6000 Blackwell, A100 (40GB/80GB), A30, L4.

India-incorporated cloud and managed hosting provider with a Noida data center. Confirmed: publishes only monthly/6-month/12-month reserved GPU pricing (no hourly option) — their own FAQ states hourly estimates require using their pricing calculator separately. GPU pricing is split across per-model URLs (e.g. one page per GPU type).

India-based data center and cloud services company with Noida and Jaipur facilities. Confirmed: their GPU pages (e.g. the H100 product page) don't publish self-service pricing at all — it's a 'Talk to Sales' / contact-for-quote model, not a public rate card. Pricing here remains an unverified placeholder since there may be no public rate to scrape in the first place.

India-based hosting and cloud infrastructure provider with a Mumbai region. Confirmed GPU catalog: A4000, L4, A5000, RTX 6000 Ada, A6000, A40, L40S, H100 PCIe, A100 80GB, H200. Publishes monthly-only pricing (no hourly billing option shown on the page).

US-based GPU cloud provider built specifically for ML training and inference. Confirmed catalog: HGX B200, H100 (SXM/PCIe), GH200, A100 (SXM/PCIe, 40GB/80GB), A6000, A10, Tesla V100, Quadro RTX 6000. Same GPU model is often listed multiple times at different prices across data center regions — pricing here reflects the cheapest available region per model.

India-based data center operator (Hiranandani Group) running 'Shakti Cloud' — hosted at the separate shakticloud.ai domain, not yotta.com. Offers GPU compute via multiple product lines: monthly 'AI Lab' GPU slices, and hourly pricing for an Azure ML Studio partnership offering (confirmed ₹264/GPU/Hr for H100, cross-validated against the AI Lab monthly rate).

India-based GPU cloud provider (Indore/Mumbai) with per-hour billing and an explicit 'per-hour, no hidden fees' positioning. Confirmed catalog: H200, H100, A100 80GB, RTX 5090 — all with real hourly pricing shown directly on the page. B200 is mentioned only in a 'reserve capacity' promotional banner, contact-only, no self-service price.

India-based GPU cloud provider with per-GPU hourly pricing (quoted in USD) plus 8x-GPU monthly bundles and managed Kubernetes (VKE) node pricing. Confirmed catalog: L4, L40S, H100 (SXM and NVL variants, same price), H200.

Global GPU cloud marketplace with per-minute billing, spot and on-demand pricing tiers. Confirmed 10-model catalog: R100, H100, B200, H200, GH200, A100, RTX PRO 6000, RTX 5090, L40S, RTX 4090 (B300/GB300/GB200 also listed but not yet verified — different card structure). Not India-incorporated.