Skip to main content
Live
Main content

Sarvam opens Token Factory with DeepSeek and Kimi models hosted in India

The IndiaAI Mission-backed startup says it already serves more than 450 billion DeepSeek V4.1 Flash tokens a day on Indian GPUs.

Jaeden Schafer
Editor in Chief · · 3 min read
Sarvam opens Token Factory with DeepSeek and Kimi models hosted in India

Indian AI startup Sarvam has opened Token Factory, an India-hosted platform that sells developers and businesses API access to its own models alongside open-weight models from DeepSeek, Moonshot AI and Z.ai. Token Factory launched November 16 and runs computation on GPU clusters in India.

Sarvam says the platform is already serving more than 450 billion DeepSeek V4.1 Flash tokens per day. The company has not said whether that figure counts input tokens, output tokens or both.

Token Factory bundles multiple models behind a single API. Other descriptions of the catalog also list Moonshot's Kimi, Alibaba models and Google's Gemma alongside Sarvam's in-house models.

Key facts

  • 01Sarvam launched Token Factory on November 16, offering many models through one API on GPU clusters in India.
  • 02The catalog includes Sarvam's own models plus open-weight models from DeepSeek, Moonshot AI and Z.ai.
  • 03Sarvam says it serves more than 450 billion DeepSeek V4.1 Flash tokens per day.
  • 04Sarvam is backed by the IndiaAI Mission, Lightspeed and Khosla Ventures, and was picked in 2025 to build India's sovereign LLM.
  • 05India's DPDP penalties reach ₹250 crore for failing to maintain reasonable security safeguards.

Sarvam has support from the government-backed IndiaAI Mission and investment from Lightspeed and Khosla Ventures. In 2025 the government chose Sarvam to develop India's sovereign large language model, and the company works with C-DAC on sovereign infrastructure.

That backing is the root of the pushback. After launch, one X user argued that Sarvam was given subsidies to build a frontier model, not to host open-source Chinese models. Co-founder Vivek Raghavan has previously warned against India exporting data while importing intelligence.

Hosting the models on Indian servers keeps inference compute in the country. There is no verified evidence that Sarvam sends customer data to China or that the hosted models contain backdoors.

India's Digital Personal Data Protection Rules, notified in November 2025, make most substantive obligations enforceable from around May 13, 2027. Penalties reach ₹250 crore for failing to maintain reasonable security safeguards.

Co-founder Pratyush Kumar said revenue from model access can help fund model development, which requires substantial capital. Sarvam plans to keep developing AI models for India and aims to become a full-stack AI company.

Related · from this week
HPE argues enterprises should own AI capacity as production workloads scale
Jaeden Schafer · 4 min read →
ShareXLinkedInEmail
AI Box

Every AI model. One chat.

The latest models from ChatGPT, Claude, Gemini, Sora, ElevenLabs — 80+ models in a single chat. Compare answers side by side. Pick the best one every time.

  • ChatGPT, Claude, Gemini, Grok, DeepSeek — in one chat
  • Generate images & video with Sora, Veo, Ideogram
  • Compare any two models side by side
  • From $8.99/mo · 80+ models, all included
Try AI Boxaibox.ai
Trusted by 3,000+ teams
Got a tip?

Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.

Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.

AI Box Daily briefingFree · Daily · No fluff

Stay ahead of everyone in AI.

The tightly edited AI news email engineers, founders, and investors actually open. One email. Every weekday. Five minutes to finish.

Loved by 10,000+ AI professionals
Free forever. Unsubscribe with one click.

The briefing read inside teams at

Keep reading

More from Business

HPE argues enterprises should own AI capacity as production workloads scale
Business

HPE argues enterprises should own AI capacity as production workloads scale

Deloitte data shows the share of firms with 40% of AI projects in production is set to double within six months, changing the cost math.

Jaeden Schafer4 min read
Nvidia logo
Business

China weighs letting ByteDance and Alibaba buy new Nvidia chips

A policy reversal would reopen one of Nvidia's largest export markets after months of Beijing steering domestic buyers toward local silicon.

Jaeden Schafer4 min read
Blue Cross Blue Shield says AI hospital coding added $942M in costs
Business

Blue Cross Blue Shield says AI hospital coding added $942M in costs

A BCBSA analysis blames AI-assisted claims tools for a two-year spending surge, with no matching change in care delivered.

Jaeden Schafer4 min read