Sarvam Launches Sarvam Token Factory: Bringing India-Hosted Inference to 8 Open Models
What happened
Sarvam has opened Sarvam Token Factory to everyone.
Performance: Fastest first token from Mumbai on most tested cases, per Sarvam’s own measurements. It serves Sarvam 105B, GLM 5.3, DeepSeek V4.1 Flash, Kimi K3.1 Pro and Gemma 4 31B from GPU (a chip built for many calculations at once, used for graphics and AI) clusters in India through one OpenAI-compatible API (the interface one piece of software uses to talk to another).
Sarvam Token Factory is an inference (running a trained model to get an answer, rather than training it) platform from Sarvam that serves its own models and open-weight (a model whose trained parameters anyone may download) (published so anyone may download the trained model) models from GPU clusters in India. Developers reach every model through one OpenAI-compatible API .
For India-based teams, this means low-latency (the delay between asking for something and getting it) tokens without sending traffic to overseas clouds. TL;DR Size: 8 models listed at launch, including Sarvam 105B (105B total, 9B active, 128K context), GLM 5.3, DeepSeek V4.1 Flash 1017, Kimi K3.1 Pro and Gemma 4 31B.
Runs on: NVIDIA B200 clusters in India. Sarvam 105B now fits on 1 B200 per replica.
It is a managed inference stack, not a single model. Sarvam first previewed it as Sarvam Inference at its Epoch event.