DeepSeek V4 Pro API pricing
DeepSeek V4 Pro by Sarvam AI costs ₹115 per 1M input tokens and ₹350 per 1M output tokens (1M in + 1M out = ₹465). Approximate USD: $1.23 in / $3.73 out (1M in + 1M out ≈ $4.95) at the RBI reference rate of ₹93.8579 per USD (2026-11-09).
- API model id
deepseekv4-pro- Input / 1M
- ₹115≈ $1.23
- Output / 1M
- ₹350≈ $3.73
- Cached input / 1M
- ₹4≈ $0.0426
- Context window
- 1.05M
- Notes
- Provisioned Throughput; enabled per organisation. Priced in INR excl. GST (as published). Approximate USD shown from the site RBI reference rate; Sarvam Token Factory USD toggle uses a fixed ₹94 per USD.
- Source
- https://docs.sarvam.ai/api/getting-started/pricing
- Provider type
- First-party (HQ: India)
- Regions
- India
- EU data residency
- no — Trust Center: managed production on Microsoft Azure Central India; LLM inference on sovereign GPU infrastructure operated by Yotta and NxtGen in India. Token Factory: prompts run on Sarvam GPUs in India and are not sent to foreign model providers. Privacy Policy also allows cross-border transfers for some processing (US/EU vendors) with SCCs.
- Currency
- INR (as published, not converted); ≈ USD at RBI reference rate ₹93.8579 per USD (2026-11-09), informational only
- Last verified
- 2026-11-12
Cost estimate
| Provider | Model | Cost |
|---|---|---|
| Sarvam AI | DeepSeek V4 Pro | ₹202.50 ≈ $2.16 |
Uses standard list prices (no caching, batch or off-peak discounts). €, CHF and ₹ rows are priced in that currency by the provider and not converted, so compare them with care. ₹ rows also show ≈ USD at the RBI reference rate of ₹93.8579 per USD (2026-11-09).
Other Sarvam AI models
- Sarvam 105B: ₹15 / ₹60
- Sarvam 105B Conversations: ₹15 / ₹60
- GLM-5.3: ₹126 / ₹390
- GLM-5.3 Flash: ₹13.5 / ₹45
- Gemma 4 31B: ₹35 / ₹90
- DeepSeek V4.1 Flash 1017: ₹14 / ₹55
- DeepSeek V4.1 Flash 1017 Interactive: ₹25 / ₹105
- Kimi K3.1 Pro: ₹200 / ₹950