India's sovereign AI inference — DPDP compliant, zero retention by default. Explore models →

Tensor Machine
India-sovereign inference

Frontier AI models.Your data never leaves India.

One OpenAI- and Anthropic-compatible API, served from Indian compute and billed in rupees.
Start buildingRead the docs
  • DPDP Act 2023
  • RBI Data Localisation
  • MeitY Guidelines
  • Zero retention by default

The frontier, priced to the paisa

ModelContextCapabilitiesInput ₹/1MOutput ₹/1M
tm/glm-5-3Z-AI1MChat · Code157.65472.96
tm/kimi-k3MoonshotAI256KChat · Code105.10420.41
tm/minimax-m3MiniMax1MChat · Code · Vision42.04210.21
tm/deepseek-v4-proDeepSeek1MChat · Code183.93367.86
tm/deepseek-v4-flashDeepSeek1MChat15.7731.53

Per million tokens, exclusive of GST. 4 more models in the full catalogue. New models land on the same endpoint — your integration doesn't change.

Full catalogue →

One line changes. Nothing else does.

Both wire formats are served — your SDK, prompts and streaming code keep working.

  • Same SDKs, same request and response shapes
  • Streaming, tool calls and JSON mode included
  • One key for every model on the rate card
from openai import OpenAI
 
client = OpenAI(
    api_key=os.environ["TENSORMACHINE_API_KEY"],
    base_url="https://edge.tensormachine.ai/v1",
)
 
client.chat.completions.create(
    model="tm/kimi-k3",
    messages=[{"role": "user", "content": "Namaste"}],
)

Built so the compliance review is short

INDIA BORDERYOUR APPEDGE · MUMBAIGPU · INDIArequestresponselogging off unlessyour org opts inresidency guardrailin the request pathno trainingpipelineThe request path. It does not reach the border.

Never.

Your prompts are not training data — there is no training pipeline, and content logging is off unless your org turns it on, enforced at the edge.

Inside.

Every request is answered from India-region compute — residency is a guardrail in the request path, not a policy clause.

Same list.

Localisation rules usually shrink your shortlist; here the frontier models and the residency requirement are the same list.

Aligned with

DPDP Act 2023
Data fiduciary controls
RBI localisation
Payments and financial data
MeitY guidelines
Governance and AI policy
SEBI
Capital-markets workloads
CERT-In
Incident reporting posture
ISO 27001 / SOC 2
Held by AWS, Mumbai region

The infrastructure certifications above are held by AWS for its Mumbai region. Tensor Machine does not itself hold these certifications yet.

Priced in rupees. Invoiced with GST.

AI infrastructure that fits an Indian budget process — not a dollar charge your finance team explains every month.

No forex, ever
Quoted and charged in INR — no conversion line on your statement.
Invoices you can file as-is
Numbered, immutable, SAC-coded, with the CGST/SGST split.
Prepaid, with hard caps
Per-team limits enforced at the edge — a runaway job stops at the cap.
Full pricing →

Tax invoice

TM/2026-27/000318

Specimen

Kimi K3

8M in · 2M out

₹1,681.62

GLM-5.3

4M in · 1M out

₹1,103.56

Subtotal₹2,785.18
CGST @ 9%₹250.67
SGST @ 9%₹250.67
Total payable₹3,286.51

SAC 998314 · Place of supply: Karnataka · Rates from the live catalogue

Regulated enterprises

Sovereign, and provably so

Deploy AI in BFSI, insurance, healthcare and government without the cross-border data-transfer review attached to it.

  • Document processing and underwriting support
  • Clinical and claims summarisation
  • Market research and filings analysis
  • Fraud detection over payments data
Talk to our team →

Builders

A key and one base URL

Start on the free tier, ship on the same endpoint, and pay in rupees when you scale. No procurement call required.

  • Coding agents in your IDE and CI
  • Multi-step agents with tool calls and audit trails
  • RAG over internal knowledge bases, incl. Indic languages
  • Customer-facing chat in 22+ Indian languages
Create a free account

Further reading

All posts →
Tensor Machine Is Live: Sovereign AI Inference, Built for Bharat

Announcement · Jan 2026

Tensor Machine Is Live: Sovereign AI Inference, Built for Bharat

After two years of building in stealth, India's first dedicated sovereign AI inference platform is live today. No prompt data leaves Indian borders. No foreign exchange markup. No DPDP grey areas. Here is what we built, and why it matters for Indian enterprises right now.

₹80 Lakhs vs ₹18 Lakhs: The True Cost of AI Inference for Indian Teams

Analysis · Dec 2025

₹80 Lakhs vs ₹18 Lakhs: The True Cost of AI Inference for Indian Teams

We modelled the total cost of AI inference for a mid-sized Indian SaaS company processing one billion tokens per month. US-based providers cost ₹80 lakhs all-in once you account for FX exposure, GST reverse charge, and compliance overhead. The same workload on Tensor Machine costs ₹18 lakhs. Here is the full breakdown.

DPDP Act 2023 and Your AI Product: What Engineering Teams Actually Need to Know

Compliance · Dec 2025

DPDP Act 2023 and Your AI Product: What Engineering Teams Actually Need to Know

India's Digital Personal Data Protection Act 2023 has direct implications for any product that processes personal data through AI inference. Most engineering teams do not know where their prompts go, who processes them, and whether they qualify as personal data under Indian law. This guide breaks it down in plain terms.

Stay ahead of India's AI curve

Get updates on DPDP compliance, new model releases, and sovereign AI. No spam, and you can unsubscribe at any time.