India pricing guide
AI agent cost in India, explained by workload.
Two projects called an ‘AI agent’ can have completely different operating responsibilities. A useful price comparison separates software licences from workflow design, integrations, data readiness, human approvals, testing, measurement, usage, and ongoing support.
Scope before software
What a credible first deployment must define.
Setup and workflow design
Price starts with the operating role: trigger, inputs, decisions, outputs, exceptions, acceptance criteria, and accountable owners.
- Number and variability of workflow paths
- Quality of existing SOPs, examples, and approval rules
- Required discovery, baseline measurement, and acceptance testing
- Single-agent scope versus coordinated multi-agent architecture
Integration and data readiness
Supported, well-documented systems cost less to connect than custom APIs, unstable exports, manual records, or broad ERP changes.
- Number of source systems, destinations, and action types
- API availability, authentication, rate limits, and sandbox access
- Record quality, field mapping, duplication, and migration needs
- Messaging, telephony, model, and infrastructure provider costs
Risk and control requirements
A read-only internal summary has a different assurance burden from an agent that contacts customers, changes records, quotes prices, or triggers irreversible actions.
- Identity, permissions, retention, encryption, and audit logs
- Human approvals, exception routes, and confidence thresholds
- Normal, failure, duplicate, unavailable, and conflict-path testing
- Compliance or contractual controls required by the workflow
Recurring operating cost
The setup quote is not the total cost of operation. Forecast recurring charges and the internal ownership that remains after launch.
- Model/API usage based on volume and task complexity
- WhatsApp, SMS, email, telephony, hosting, and storage charges
- Monitoring, corrections, policy updates, and managed support
- Human review capacity for exceptions and high-impact decisions
Buying model comparison
Compare total operating cost—not isolated price labels.
A low monthly licence can be economical when internal implementation is simple. A managed quote is justified only when it covers material delivery, integration, control, and operating work that your team would otherwise own.
Upfront spend
Low-code / SaaS routeUsually lower, excluding internal implementation time and custom work.
Managed Shoris route₹25,000–₹40,000 + applicable GST for decision-ready discovery when required; From ₹1,50,000 + applicable GST for deployment. Most scoped deployments: ₹1.75–₹2.5 lakh + applicable GST.
Implementation
Low-code / SaaS routeInternal team configures workflows, prompts, connectors, QA, and handoffs.
Managed Shoris routeWorkflow, integration, controls, testing, acceptance, and measurement are defined in the engagement scope.
Recurring spend
Low-code / SaaS routeLicence tier, usage, channels, and your own maintenance time.
Managed Shoris routeUsage, channels, and infrastructure remain separate. Managed Agent Care is From ₹25,000/month + applicable GST; most active deployments: ₹35,000–₹65,000/month + applicable gst.
Best fit
Low-code / SaaS routeStandard workflow, supported connectors, low risk, and capable internal owner.
Managed Shoris routeCross-system or custom workflow requiring accountable implementation and governance.
Main pricing trap
Low-code / SaaS routeComparing licence price without internal build and maintenance effort.
Managed Shoris routeTreating a starting price as fixed before systems, data, actions, and exceptions are known.
Fit check
Know when to deploy—and when to wait.
The right first workload is repetitive, measurable, digitally accessible, and has a named human exception owner.
Good first scope
- The workload has measurable volume, delay, and paid effort
- Systems, actions, and exception ownership can be scoped
- The buyer can fund setup plus realistic recurring costs
- The team will measure the first workload before expansion
Not ready yet
- Budget is based only on the cheapest software licence
- Required data cleanup and custom integration are assumed free
- The decision depends on guaranteed ROI or headcount removal
- No owner can provide access, policies, testing, or exception review
Buyer objections
Questions to resolve before signing.
Is ₹1.5 lakh the final price for every single agent?
No. It is the public starting price. The final quote depends on workflow paths, integrations, data readiness, controls, channels, acceptance criteria, and recurring support. The scope should state each assumption and exclusion.
When is the Workflow Blueprint credited?
Credited in full when the deployment scope is signed within 30 calendar days of blueprint delivery. The blueprint still has standalone value: it defines the workflow, access, controls, KPI baseline, exclusions, and acceptance criteria before engineering begins.
Why is a managed agent more expensive than chatbot software?
They are different purchases. Software provides product access; a managed deployment can include workflow design, integration, permissions, guardrails, testing, measurement, and operating ownership. Choose software when your team can responsibly own those layers.
Are API, WhatsApp, infrastructure, and support included?
Model/API usage, messaging or telephony charges, infrastructure, data cleanup, custom ERP work, and work outside the signed acceptance criteria are scoped separately.
Can the quote be justified using guaranteed savings?
No. Capacity returned and setup-cost recovery are estimates based on supplied workload bands and assumptions. The proposal should show the baseline, uncertainty, recurring costs, and what will be measured after launch.
When should we buy a department rollout?
After at least two coordinated workloads, shared systems, governance needs, and an expansion case are understood. The public department price is a floor, not a promise that any number of agents is included.
Continue the evaluation
Continue your deployment evaluation.
No-pressure scope check
Test the economics before requesting a quote.
The Agent ROI Audit compares suitability, capacity returned, answer confidence, and setup-cost recovery. It is designed to reject weak cases instead of converting every visitor into a proposal.
Start the Agent ROI Audit