News

Sarvam AI Opens Token Factory, Challenges OpenRouter in India

Sarvam AI has opened its Token Factory platform to developers and businesses, offering multiple AI models through one API with rupee-based pricing and computing infrastructure hosted in India.

Written By : Akshita Pidiha
Reviewed By : Vijay Laxmi

Indian artificial intelligence startup Sarvam AI has opened its Token Factory platform to developers and businesses. It is allowing them to access multiple AI models through a single API. The service includes Sarvam's own models alongside global offerings such as DeepSeek, Kimi and Google's Gemma.

The platform aims to simplify AI deployment for businesses. Users can switch between supported models without building separate infrastructure or managing individual service connections.

Token Factory uses an OpenAI-compatible API. Developers already working with OpenAI's API can connect their applications by changing the endpoint and selecting a supported model. A single Sarvam API key, software development kit and credit balance can be used across the platform.

Its launch catalogue includes Sarvam 105B, GLM 5.3, GLM 5.3 Flash, DeepSeek V4.1 Flash, DeepSeek V4 Pro, Kimi K3, Kimi K3.1 Pro and Google's Gemma 4 31B. The wider catalogue also features Qwen 3.8 27B, an embedding model and Prism, Sarvam's decision model.

India-Based GPU Infrastructure

Sarvam hosts the supported models on GPU clusters located in India instead of routing requests to overseas infrastructure operated by model providers. The company says this arrangement can improve response times and give businesses better control over data processing locations.

Sarvam has also disclosed that it operates India's largest cluster of NVIDIA B200 GPUs. It is expanding its computing capacity to support additional models and rising demand. The paid service follows a usage-based pricing model, with charges calculated according to input and output tokens. All prices are listed in Indian rupees and exclude GST.

Sarvam 105B carries introductory rates of Rs. 10 per million input tokens and Rs. 25 per million output tokens. For DeepSeek V4.1 Flash, Standard pricing stands at Rs. 14 per million input tokens and Rs. 55 per million output tokens. Its Interactive option costs Rs 25 and Rs. 105, respectively. Kimi K3.1 Pro is priced at Rs. 200 per million input tokens and Rs. 950 per million output tokens. Pricing for other models varies across the catalogue.

Also Read: Sarvam Becomes India's Newest AI Unicorn With $234 Million Funding Led by HCLTech

Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

Trezor Data Breach Exposes 27,000 Orders as Scam Calls Begin

Crypto News Today: Tether Capital Invests in Nomad, Binance Hacker Moves Funds, PI Hits All-Time Low

Ethereum Price Recovers Above USD 1,800 as Goldman Sachs Expands ETF Holdings Despite Outflows

Ex-Robinhood Engineer Pleads Guilty to USD 125K Crypto Fraud

Cardano Community Rejects Copper Tokenization Plan Over Treasury Concerns