SOVEREIGN MODEL PORTFOLIO
One sovereign perimeter. A family of models built for India’s institutions.
Taksha, our coding family, is serving developers today — benchmarked against frontier models on identical execution-graded tasks, with the results published including the ones we lose. The families below apply that same discipline to reasoning, regulated domains and the edge.
THE STACK
One sovereign stack — from foundation engines to specialist minds.
Every family stands on a foundation we operate, and everything ships inside the same trust layer — not beside it. The saffron slab is the lineage trained from scratch in India.
The trust perimeter — every request crosses it
Gateway · Veil · Kavach · TrustFabric · Custodian · Governance
Our sovereign AI engineer, in developers’ hands now — with an independent referee that checks its work and says so when a claim doesn’t verify.
GA · now
- Taksha ProThe everyday default — edits, runs, verifies and finishes. 45/45 agentic, zero phantom completions.
- Taksha Max45/45 on the graded agentic battery. 160K context, dedicated node, provisioned on request.
- Taksha AdhunikLegacy modernization under a migration doctrine.
- Taksha ParikshaTesting that bites — proves the suite goes red before claiming a pass.
- Taksha YantraFor teams building agentic AI — frameworks, tool protocols, evals. Alpha · Target GA Q4 2026.
- Taksha KritiInterfaces that are wired, not just written — +15 pts over the base across 8 UI classes.
Sovereign engines
Pre-trained engines — operated & fine-tuned by us, served from India or inside your perimeter.
Indigenous
Trained from scratch in India — Tatva Indic Nano (first checkpoint trained, 32k Indic tokenizer); more in training.
EVERY MODEL EARNS ITS MARK
We publish status as plainly as we publish benchmarks. Taksha is generally available today. Every other family is moving through the same proving ground Taksha graduated from — execution-graded testing on our own infrastructure, hardened against real workloads with design partners and early-access customers. A model earns its GA mark here; it is never granted. So by the time a family reaches your environment, the testing is already behind it — your deployment is the destination, not the experiment.
HOW THE PORTFOLIO IS NAMED
Two axes, one vocabulary — across every family.
CAPABILITY TIER
Nano → Fast → Pro → Max → Ultra
The same five words mean the same five things everywhere. Nano runs on a phone; Fast is the everyday driver; Pro is the fine-tuned flagship; Max brings frontier-class reasoning and long context; Ultra is reserved for dedicated, customer-funded capacity. Learn it once and you can read any model in the portfolio.
PURPOSE NAME
Yantra · Adhunik · Pariksha · Kriti · Kuber · Seva · Astra · Kisan · Sanjaya
A Sanskrit name states what a model is for, not how big it is. Taksha Adhunik modernizes legacy estates; Taksha Pariksha writes tests that bite; Kuber knows BFSI; Kisan speaks to the field. Any model reads as family · purpose · tier — for example Taksha Pariksha Pro or Tatva Kisan Nano.
Tiers listed on each card below are the ones planned for that model. A tier appears in the product only once it exists and has been measured — the naming system is a map of intent, not a claim of availability.
THE MANTHAN LINE · DOMAIN INTELLIGENCE
For the work your regulator asks about.
Hosted reasoning models with a chat surface and an API — general, or shaped to a regulated domain. Served in India or inside your own network.
Manthan मंथन
General reasoning
Fast · Pro · Max · Ultra
The sovereign assistant for everyday enterprise work — the class of tool your teams already paste confidential documents into, except it answers from inside your perimeter.
- Live now at chat.agentanywhere.ai — sign in and try it
- Chat and API surface, OpenAI-compatible
- Served in India or inside your VPC
- No prompt or document leaves the perimeter
Kuber कुबेर
BFSI
Pro · Max
Banking, insurance and capital markets cannot send customer data to a foreign cloud — and increasingly cannot justify it to a regulator. Kuber is built for that constraint first.
- Regulatory and circular comprehension
- Credit, claims and compliance workflows
- Audit trail on every answer
Seva सेवा
BPO & service operations
Fast · Pro
India runs the world's service desks under client contracts that bind where data may live. Seva is the assistant those contracts can actually accommodate.
- Agent assist and ticket comprehension
- Quality review of interactions at scale
- Deployable per-client, per-jurisdiction
Astra अस्त्र
Defence & strategic
Pro · Max · Ultra
For institutions where a network egress is a disqualifier: fully air-gapped deployment, no telemetry, no vendor dependency across a border.
- Air-gapped, no outbound connectivity required
- Document and imagery analysis under your control
- Engagement under formal procurement only
THE TATVA LINE · INTELLIGENCE AT THE EDGE
Where the network isn’t.
A data centre is an assumption. Most of India — the field, the border, the factory floor — does not make it. The Tatva line runs on the device, offline, in the language spoken there.
Tatva तत्त्व
On-device reasoning
Nano · Fast · Pro
Intelligence that runs where the network does not. Modest hardware, no cloud round-trip, no connectivity assumption.
- Runs on-device, fully offline
- Indic-language capable
- The same sovereign story, without a data centre
Tatva Kisan किसान
Agriculture
Nano · Fast
Advisory for the farmer standing in a field with one bar of signal — in their language, on a device they already own.
- Crop, input and advisory guidance offline
- Indic languages, voice-first by intent
- Built for the connectivity India actually has
Tatva Indic Nano नैनो
Indic, from scratch
Nano
The sovereignty maximalist position: small, trained from scratch, lineage fully disclosed and fully ours — nothing inherited, nothing to explain away.
- Trained from scratch, not adapted
- Auditable provenance end to end
- Indic-first by construction
Tatva Drishti दृष्टि
Geospatial detection
Nano · Fast
Object detection on satellite and aerial imagery, running at the edge — for monitoring that cannot depend on someone else's cloud.
- Detection on satellite and aerial imagery
- Edge inference, no imagery egress
- Infrastructure, agriculture and strategic monitoring
Tatva Sanjaya सञ्जय
Telemetry & anomaly detection — defence, space, energy, industrial
Nano · Fast
The seer who watched the distant war and narrated it live to the blind king. Sanjaya watches thousands of live sensor channels at once — a spacecraft from its ground station, a grid, a plant, a network — catching anomalies, warning of a breach before the red-line is crossed, and naming the channel at fault.
- Detects anomalies and predicts a breach before the red-line — with early warning validated on real spacecraft telemetry (ESA anomaly dataset)
- Multivariate time-series sentinel: point, contextual and temporal anomalies, limit monitoring, root-cause channel attribution and recommended operator actions
- One capability across sectors: defence & space (ground-segment, mission telemetry), power grids & energy, industrial plants and SCADA, and telecom networks
- Trained from scratch — no pre-trained base, lineage disclosed end to end. A compact proof model today, running on-prem / air-gap-friendly with telemetry never leaving the room
WHY THIS PORTFOLIO
The discipline is the product.
Sovereign by construction
Models we own and operate, served from India or inside your perimeter. Data residency as a network fact, not a contractual promise.
Verified, not asserted
The independent referee we ship with Taksha — a check the model never sees — is a portfolio principle, not a coding feature. It travels to every family.
Measured in the open
We publish the benchmark table including the cells we lose. Ask any other vendor for theirs; the silence is the comparison.
Indian by intent
Indic languages, Indian regulation, Indian connectivity and Indian procurement — designed for, not translated into.
THE TRUST LAYER
Every model here ships inside the same trust stack — not beside it.
A sovereign model is only as sovereign as the path a request takes to reach it. The portfolio serves behind the AgentAnywhere trust layer — live today, governing production traffic.
Kavach
Inline AI-security shield
कवच — shield. Sits in front of any OpenAI-compatible model and enforces security on every request and response: ensemble threat detection inbound, secret redaction outbound, PII masking via Veil, and an audit trail that stores metadata — never raw values.
THE TRUST LAYER · IN MOTION
Your data’s bodyguards — every request, every time.
Watch one request make the trip. Nothing reaches the model unguarded — policy checked at the door, personal data masked before the model ever sees it, threats stopped at the shield. And nothing leaves unsigned: every answer carries a receipt, and every request — allowed or blocked — lands in the evidence pack your auditor reads.
Now showing: a clean request, end to end
Incoming request
“KYC update for PAN ABCDE1234F⟨PAN_1⟩, A/C 0000 1111 2222⟨ACC_1⟩”
“KYC update for PAN ABCDE1234F, A/C 0000 1111 2222”
- policy: allowed · logged
ABCDE1234F ⟨PAN_1⟩ 0000 1111 2222 ⟨ACC_1⟩
- clean request: threat scan cleaninjected prompt: blocked
- processing inside the perimeter
- receipt tf_9f42…c1a ✓ signed
- evidence pack appended · SOC 2 · ISO 27001 · RBI-aligned
Each guard is a real product, live today: Gateway · Veil · Custodian & TrustFabric · Governance.
FAQ
Frequently asked questions.
- Which of these models can I use today?
- Taksha, the coding family, is available now — invite-only during the preview. Every other family is in private preview, early access, or a design partner program with its target GA month stated on its card — none of them serve production traffic yet. We would rather say that plainly than have you discover it during your evaluation.
- When will Manthan, Kuber, Seva or the Tatva line be available?
- Each ships when it passes the same execution-graded gate Taksha passed, not on a marketing date. Early-access participants are told the honest status at every step, including delays. Register interest and we will bring you specifics rather than a countdown.
- Can these run inside our own network or air-gapped?
- That is the design intent for the entire portfolio, and it is already delivered for Taksha. The Astra line assumes air-gapped operation as a starting condition rather than an option.
- What does sovereignty mean here, concretely?
- Models we own and operate rather than rent, served on infrastructure in your jurisdiction or inside your perimeter, with no requirement for data to cross a border for the product to function.
- How do you evaluate models before claiming anything?
- Execution-graded tasks with identical graders for every model, ours and the frontier ones. Tests are mutation-checked, servers are actually called, and claims a model makes about its own work are verified by a command it never sees. We publish the failures alongside the wins.
Tell us which of these you need first.
Early access is a conversation, not a waitlist form: we will tell you the honest status, what it can and cannot do today, and what a pilot would actually prove. Taksha is available now if you want to see the discipline before the roadmap.