AI infrastructure Model or AI service
Sarvam’s Token Factory serves language models from GPU clusters in India
Sarvam introduced Token Factory as a hosted service for its models and selected open-weight models. [1]
What the model does
Its inference stack runs the supported models on GPU clusters in India and exposes a hosted interface for client applications. [1]
Why it matters
A hosted service can handle inference infrastructure for a team that wants to test a supported model without running the weights on its own servers. [1]
Key dates
- Sarvam introduces Token Factory [1]
Sources and verification
Catalog evidence checked How we verify records
- Primary source Introducing Sarvam Token Factory | Sarvam AI
- Independent source Sarvam Opens India-Hosted Token Factory for Open-Weight Models – Unite.AI
Related profiles
AI infrastructure
NVIDIA Rubin HPC
NVIDIA
It is designed to accelerate high-performance computing workloads in compatible systems.
AI infrastructure
Intel Crescent Island (D270) GPU
Intel
It is a data-center GPU design for workloads described in Intel's reported technical presentations.
AI infrastructure
AMD Instinct MI450X
AMD
The accelerator is designed for data-center AI compute in compatible infrastructure.