Lexlegis.ai vs Manupatra: how they compare in 2026
Lexlegis.ai and Manupatra both put generative AI over Indian case law and statutes for lawyers, tax professionals and in house counsel. Lexlegis.ai leads on where it can run and who vouches for it. It publishes five deployment modes, from Indian cloud to air gapped installations, holds ISO 27001 with SOC 2 reports on request, and names KPMG and the Government of India among customers. Manupatra's lead is in its contract and its corpus. Its agreement bars training on uploaded material, including anonymized use, without separate written consent, and requires deletion within 24 hours with breach notice inside the same window. Its own flags mark whether a judgment is still good law. Manupatra has covered Indian courts since 2000, across more than 10 million documents. Both claim their AI eliminates hallucinations.
Which one fits
Choose Lexlegis.ai if
- You need the AI inside your own walls. Lexlegis.ai publishes five deployment modes, from multi tenant cloud in India through a single tenant sovereign Indian cloud and your own AWS, Azure or Google estate to on premises and air gapped installations, with region locking and customer managed keys above the entry mode.
- Your procurement team wants reports. Lexlegis.ai states ISO 27001 and SOC 2 Type 2 with audits by government empaneled assessors, and its trust center offers the certificate, the SOC 2 report and a penetration test report through a request flow.
- You want an agent for multi step legal work. Lexlegis.ai's MIRA selects from 215 production skills, orders them into a plan and checks its reasoning before output, and every factual claim in an output is traced to a source, with broken traces flagged.
Choose Manupatra if
- You need to know whether a judgment is still good law. Manupatra marks statutes and judgments with its own blue, yellow and red flags and effective date highlights, and its AI Search can show whether a judgment has been overruled, distinguished or followed.
- You want strict handling of uploaded case files in the contract. Manupatra's agreement bars training on subscriber content, including anonymized or aggregated use, without standalone written consent, requires deletion within 24 hours including AI memory and logs, and indemnifies you for confidentiality breaches.
- You want to price AI by the task. Manupatra publishes AI Summary at ₹5,000 a year per court level and a credit rate card for its toolkit, from 3 credits a page for summaries to 50 a query for drafting, at ₹1.25 a credit.
At a glance
All 15 axes, side by side
The same grid applied to every vendor in the index, graded from public sources. Hover a grade to see what the letter means on that axis.
AI Centrality
How much of the product is actually AI. Whether the machine learning is the mechanism the buyer is paying for or a feature layered onto conventional software, and whether the vendor is specific about which is which.
Every product sold is model work, and the company describes itself as AI first. Legal AID is three generative capabilities. Ask answers legal questions in the IRAC format with sources returned and explained, Interact reads uploaded contracts, orders and notices and answers across them, and Draft produces contracts, submissions, arguments and authority bundles. MIRA, on the same foundation, selects from a library of 215 production skills, orders them into a plan and checks its own reasoning against a verification layer. The agreement defines the Service as the platform including Ask, Interact, Draft and MIRA, so the model products are the contracted thing. The vendor states it trained its own model on its legal collection, and its published model layer describes signed checkpoints, guardrails for each tenant and an inference audit trail. The asset outside the models is real: a curated closed collection of Indian laws, circulars and judgments built over decades. No product without AI is sold separately, and the collection feeds the assistants rather than being sold as a search subscription.
A substantial, named AI layer sits on a Manupatra research platform that ran for 25 years without it. AI Search launched in December 2025, is described as powered by semantic AI and large language models, and finds judgments by legal concept rather than keyword. AI Summary summarizes judgments and AI Compare sets them side by side. The ManuWorks toolkit adds OCR, summaries, translation, document questions, case timelines, comparison and drafting, each with its own credit rate. The site is headlined India's trusted Legal Research partner, Powered by AI. Manupatra still treats AI as one search mode among several. It tells users to use MANU SEARCH, LEGAL SEARCH and CITATION SEARCH for exact or Boolean queries and keeps AI Search for queries by meaning. The AI is paid for separately through a credit wallet, so a subscriber can use the platform without it. The core asset is the edited corpus and its citation and treatment links, which predate the models.
Citation Accuracy and Hallucination Disclosure
Whether the vendor publishes measured accuracy on citations and assertions, grounds output to primary sources, and says plainly what its system does when it does not know. Legal has a documented public record of fabricated citations reaching filed briefs, so an untested claim of accuracy is not evidence.
Real architectural controls sit under an accuracy claim that is never measured. Output comes in the IRAC format with a list of the sources relied on, each openable and downloadable, with a summary and a statement of why it was pertinent. Retrieval is from a closed curated database of laws, circulars and judgments rather than the open web, in the order retrieve, reason, verify. The model layer names output hallucination detection, a meta reasoning gate, guardrails for each tenant and signed model checkpoints. The vendor states that every factual claim in every output is traced back to its source and that a claim whose trace is broken is flagged, with no exceptions. MIRA is described as checking its own reasoning against a layer built to catch its mistakes. The vendor markets research without hallucination and grounded, hallucination free legal intelligence. No accuracy figure, error rate, hallucination rate, test set, benchmark or evaluation result is published for any of it, so an absolute claim stands without measurement. Its trust center lists items titled A note on Hallucinations and AI Flow for Ask.
Manupatra grounds its AI in its own edited corpus and shows where it is weak, but publishes no accuracy figure. Its AI brochure says AI Search runs on a proprietary, edited corpus rather than public web data. Results come as the ten most relevant judgments, broken into Facts, Issues, Arguments, Reasoning, Ratio and Summary. Hovering on Reference confirms the source, and clicking Citation opens the full judgment. The platform also flags overruled, distinguished and conflicting judgments, with blue, yellow and red flags for whether authority is still good. The brochure claims the edited base eliminates AI hallucinations. That claim sits alongside real controls. Manupatra also publishes where AI Search is weak. A section titled When NOT to Use AI Search warns against exact queries, party names, citations, exact statutory wording and verbatim quotes, and routes them to MANU SEARCH, LEGAL SEARCH and CITATION SEARCH. Another page gives nine examples of well and poorly framed queries. What is missing is measurement. No accuracy figure, test set, error rate or evaluation is published for AI Search, AI Summary or AI Compare, and nothing backs the claim that hallucinations are eliminated. Nothing says the system reports when it finds no support. It always returns ten results, so whether the tenth is relevant or merely nearest is unanswered. Clause 11.12 of the agreement says judgment text is computer generated and must be checked against the certified copy. Clauses 15.1 to 15.3 disclaim accuracy and put checks on the user.
Autonomy and Oversight Model
What the system decides on its own, what a lawyer must approve, and whether the vendor documents where the review point sits. A tool that drafts under review and a tool that files without one are different products and different risks.
The oversight position is contractual and an automated checkpoint is described, but the point at which the agent acts alone is not stated. Clause 8 of the terms, headed AI outputs and professional responsibility, says outputs support legal or compliance work and are not legal advice. The customer must review and validate all outputs before relying on, filing or acting on them, no output creates a lawyer client relationship with the vendor, and the Service does not replace the customer's professional judgment. Clause 7 adds that the customer will not use outputs to falsely attribute authority or to deceive any court or tribunal. On the product side, the model layer publishes a meta reasoning gate and output hallucination detection through which every input and output passes. The vendor states that failure of any layer opens a ticket rather than a breach, and MIRA is described as verifying its own reasoning before producing output. Nothing states what MIRA does without a person, when a skill chain completes unattended, whether any step needs human approval, what a user sees when the gate rejects a claim, or whether autonomy is configurable.
Clear guidance covers when not to use the AI, but Manupatra sets no thresholds or review step. Its AI brochure has a section titled When NOT to Use AI Search. It tells users to avoid AI Search for exact or literal queries, party names, case citations, exact statutory wording and verbatim quotes. It says AI Search does not follow Boolean logic and sends those queries to MANU SEARCH, LEGAL SEARCH and CITATION SEARCH. Users can check sources: hovering on Reference confirms the source, clicking Citation opens the full judgment, and a chatbot handles follow up questions. The agreement backs this up. Clause 15.3 makes the user responsible for checks, and clause 11.12 requires verification against the certified copy. What is missing: no confidence score, no threshold, nothing on what the system does with a query it handles poorly, and no review step before output. Nothing covers oversight of the ManuWorks AI Drafter, which writes legal drafts at 50 credits a query. The published limit says which queries suit AI Search, not what its output may not be used for.
Operational and Outcome Evidence
Named, dated evidence that the product works in production at real firms or legal departments. Case studies with figures and identified customers count. Unattributed testimonials and launch announcements do not.
A named customer roster with institutional weight, and no outcome measured or attributed. The vendor's homepage names KPMG, Dhruva Advisors, Greaves Cotton, Thermax, K Raheja Corp, Aurtus, MyGate, Artha Energy and the Government of India, and the vendor states more than 100 enterprises. Public milestones are dated and checkable: the platform was shown at NVIDIA GTC 2026 in March 2026, where the company announced a preview beta in the United States, and the skill library is quantified at 215 production skills. No case study is published, no customer is quoted by name and role, no deployment is described, and no outcome is measured. There is no time saved, volume processed, adoption figure, or accuracy or quality result from any customer. The efficiency claims that appear are unattributed and unquantified: research memos in hours rather than days and authority bundles built automatically. The record shows who is said to use it and nothing about what it did for them.
Scale is claimed by Manupatra, with no customer named and no outcome measured. It reports more than 150,000 active users, more than 10 million documents and judgments, 25 years in operation, a team of more than 175, and offices in India and Bangladesh. Its About page lists five kinds of client: legal professionals and firms, in house legal teams at companies and banks, academic institutions, government and regulators, and students and researchers across Asia. Categories are not customers. No law firm, company, court, tribunal or university is named, and there is no client list, case study, testimonial or logo. No outcome is measured: nothing on time saved, research hours or accuracy, and the AI brochure's benefits describe features rather than results. The 150,000 figure has no date or source and appears as a counter on the product page and in the About page metadata.
Privilege and Confidentiality Posture
How client confidences are handled: attorney client privilege and work product treatment, segregation of one client matter from another, whether client data trains any model, and what the vendor commits to in writing rather than in marketing.
Substantive commitments are published, most of them in the contract, and privilege is not addressed. Clause 10 of the terms binds each party to hold the other's confidential information in confidence and use it only for the agreement. It names Customer Content as the customer's confidential information and survives termination for five years, or indefinitely for trade secrets. Clause 5 leaves all rights in customer content with the customer and licenses the vendor to process it solely to provide the Service. The privacy policy says content data is accessible only to the account it belongs to. The trust page publishes matter level isolation on multi tenant SaaS, alongside custom roles, single sign on required on enterprise, and session recording on enterprise. Customer managed keys with hardware security module storage are available on the other modes, and retention can be set to zero on enterprise and the on premises and air gapped modes. The words privilege and work product appear nowhere in the terms, the privacy policy or the trust pages. Clause 8's statement that no lawyer client relationship arises with the vendor addresses the vendor's own position, not the customer's privilege.
Manupatra's agreement makes precise confidentiality commitments, but never mentions privilege. Clause 4.1 defines Confidential Information to include documents, data, personal data, case files, names, facts, legal strategies and client information uploaded through the site, the AI tools or any related service. Clause 4.2 requires it to be kept strictly confidential and used only to provide the service requested. It may not be copied, stored, analyzed, mined or shared for any other purpose, and only staff who need it, under equal duties, may access it. Clause 4.3 bars disclosure to any third party, including affiliates, subcontractors, AI engines and cloud providers, without express written consent. Clause 4.7 requires permanent deletion of all such information, including copies, backups, AI memory, logs and caches, right after the request is met and within 24 hours. Written certification is available on request. Clause 5.8 bars staff from viewing Subscriber Content or AI outputs except to fix a technical issue the Subscriber raised, with access logged and kept to a minimum. Clause 4.8 makes these duties survive termination indefinitely. Training requires consent. What is missing is privilege. Neither privilege, professional secrecy nor work product is named, and nothing covers privileged material if disclosure is compelled. Listing case files and legal strategies comes close, but is not the same.
UPL and Professional Responsibility Posture
Whether the vendor is clear that it supplies a tool rather than legal advice, who its audience is, and how it addresses unauthorized practice of law, competence and supervision duties, and jurisdiction limits. ABA Formal Opinion 512 is the reference point. Where the advice line is not the duty a product raises, the axis is read through the nearest professional duty it does raise: judicial conduct rules and the reviewing duty for products sold only to courts, and the duty to bill for time actually spent for products that draft time entries.
The boundary is drawn in the agreement and backed by an access gate, and no governing professional rule is named. Clause 8 states that outputs support legal or compliance work and are not legal advice, that the customer must review and validate outputs before relying on, filing or acting on them, that no output creates a lawyer client relationship with the vendor, and that professional judgment is not replaced. Clause 7 prohibits using outputs to falsely attribute authority or to deceive any court or tribunal. Clause 2 restricts paid plans to legal or compliance professionals or people in a role requiring legal research and drafting tools, which is an access control rather than a disclaimer and answers the unauthorized practice question directly. The vendor also states that every skill is designed, tested and validated by senior lawyers. No Bar Council of India rule, provision of the Advocates Act, state bar council guidance or court practice direction is cited anywhere. The Supreme Court of India has publicly taken up AI generated authority in filings.
Clause 16.1 of Manupatra's agreement, headed No Legal Advice, says material on the site is not legal advice and creates no relationship of client and advocate. The disclaimer on the legal page says Manupatra does not warrant its forms and creates no relationship of lawyer and client by providing them. It adds that forms are a guide to be adapted at the user's own risk. Clause 15.3 makes the user responsible for checks on the accuracy of input and output. Clause 11.12 requires users to verify judgment text against the certified copy, a duty rather than a disclaimer. No conduct rule is named, Bar Council of India guidance is not mentioned, and nothing asks subscribers to use the product consistently with their professional duties. In July 2026 the Supreme Court of India held that citing unverified AI generated precedent is professional misconduct and directed India's Bar Council to frame norms. Manupatra publishes that ruling as news on its own updates service, but says nothing about what its AI features mean for that duty.
AI Governance and Bias Disclosure
Published governance over model behavior: who owns it inside the vendor, what is tested before release, and what is disclosed about disparate output across matter types, parties, or populations.
Technical controls over the model are published in detail, and no governance structure is described. Two items come close to governance. Guardrails can be set for each tenant, which puts configuration of model behavior in the customer's hands, and the vendor states that every skill is designed, tested and validated by senior lawyers before it ships. The trust center also lists a code of ethics, anti bribery and anti modern slavery items, and a product security entry titled A note on Hallucinations. What is missing is governance of the AI itself. No AI governance framework or standard is published, and no ISO/IEC 42001 or equivalent is claimed, although five security and privacy frameworks are claimed precisely. Nobody is named as accountable for model behavior. No evaluation protocol before release, benchmark, red team exercise or published result exists for any model or skill. No bias, fairness or representativeness disclosure appears, for a product whose source material is Indian judicial and tax material.
Responsible AI gets one paragraph from Manupatra and nothing behind it. The legal page's Use of AI section says Manupatra uses AI responsibly. It lists data privacy, bias mitigation, transparency, human oversight, accountability, user consent and regulatory collaboration as priorities, to maximize benefits and minimize risks. That is the whole of it. No framework is adopted, no owner or committee is named, no governance process or bias testing is described, and no results are published. Nothing says how human oversight works, beyond guidance to users. Bias is named as a priority and never returned to, on a product that ranks and summarizes court decisions, where a tilt in which precedents surface would be invisible to the user. The AI section of the agreement is a strong data handling commitment, but the responsible AI paragraph is not tied to it. The Grievance Redressal Officer named on the same page is a statutory appointment under India's IT Act for complaints, not an AI owner.
AI Safety and Data Stewardship
Retention, deletion, access control, and what happens to prompts and documents after they are processed. Whether the vendor states its subprocessors and its incident practice, or leaves the buyer to assume.
Published policy covers the ground in the agreement and a structured trust architecture, with one gap. Certifications are stated precisely: ISO/IEC 27001 to the 2022 revision, SOC 2 Type 2 and ISO/IEC 20000-1:2018, with audits by CERT-In empanelled assessors. Encryption is TLS 1.3 in transit and AES 256 at rest, with customer managed keys and hardware security module storage on the modes other than SaaS. Access control covers SAML 2.0, OIDC and SCIM 2.0, single sign on required on enterprise, custom roles, multi factor authentication through the customer's identity provider, and access logs. Network controls include private cloud isolation, private link, a firewall and IP allow listing. The model layer adds prompt injection detection and signed checkpoints against model poisoning. Customer content is excluded from shared training, retention can be reduced to zero on enterprise and the isolated modes, and data is exportable at any time and returned or destroyed on termination with attestation. A disclosure program commits to acknowledgment within 24 hours and triage within 72. The gap: the privacy policy says a current subprocessor list is published at the trust center, and none appears there.
Most of the ground is covered in specific terms by Manupatra's agreement, which names no subprocessors. Clause 4.4 requires industry standard technical and organizational security. Clause 4.5 commits to written notice within 24 hours of any actual or suspected unauthorized access, loss or breach, with full details and impact. Manupatra must also contain, investigate and fix it at its own cost, and cooperate in any legal or regulatory action. Clause 4.6 makes Manupatra the data processor, with the Subscriber as data fiduciary under the Digital Personal Data Protection Act 2023, and commits to the Information Technology Act 2000. Clause 4.7 requires deletion within 24 hours with written certification. Clause 5.3 requires any storage to be temporary, encrypted and deleted after use. The privacy policy adds encryption, secure servers and firewalls, access controls, and staff confidentiality agreements and training, all regularly reviewed, and states compliance with the SPDI Rules 2011. No processor is named. Clause 4.3 bars disclosure to any third party, including subcontractors, AI engines and cloud providers, without express written consent. But the privacy policy says third parties handle payments, email and hosting, and that products may connect to Dropbox and Google Drive, so processors exist and none is named. No certification backs any of this.
AI Liability and Recourse
What the vendor stands behind contractually when its output is wrong. Indemnities, caps, carve outs, insurance, and whether any of it is published or only reachable through a negotiated agreement.
A warranty, a service level, insurance and a cap are published, and no indemnity runs to the customer. Clause 11 warrants that the Service will be provided with reasonable care and skill. Clause 9 sets a 99.5 percent monthly uptime target on the SaaS mode, with bespoke service levels for enterprise and other modes in the order form. Its trust center lists cyber insurance and professional indemnity insurance. Clause 12 caps liability at fees paid in the twelve months before the claim and mutually excludes indirect, incidental, consequential and punitive damages, while preserving liability for gross negligence, willful misconduct and anything that cannot be limited by law. Clause 13 allows termination on 30 days uncured material breach, with customer content returned or destroyed at the customer's election. No indemnity runs to the customer: no intellectual property indemnity and no defense obligation. The warranty covers the manner of provision, not the output. Clause 11 states that outputs are not warranted to be accurate, complete, error free or fit for purpose, so the customer bears the loss when an output is wrong.
Warranty, liability and indemnity are covered in detail in the agreement, and the balance favors Manupatra. Clauses 25.1 and 25.2 provide the information as is and disclaim all warranties, including merchantability, fitness for purpose and noninfringement. Clauses 15.1 and 15.2 disclaim any promise that the service will be correct, accurate, reliable, uninterrupted or timely. Clauses 25.3 and 25.4 exclude liability in contract or tort for any loss linked to the service, even where caused by Manupatra's own negligence. Clause 25.5(i) excludes liability for any claim about the subscriber's failure to do legal research properly, or any decision made in reliance on the Data. Clause 26.1 caps liability at the charges paid in the subscription period, but it is drafted for copyright claims on the data, not as a general cap. The agreement also runs the other way. Clauses 21.2 and 26.3 require Manupatra to indemnify and defend the Subscriber, its clients, staff and affiliates. That covers breach of confidentiality, misuse of Subscriber Content, unauthorized access or data breach, negligence, willful misconduct or breach of law by Manupatra, and intellectual property infringement. Clause 5.9 adds immediate termination, indemnity and injunctive relief for misuse of Subscriber Content, and clause 12.1 a prorated refund. Insurance is not addressed.
Practice Systems Integration Depth
How deeply the product reaches into the systems legal work already lives in: document management such as iManage and NetDocuments, Word and Outlook, contract lifecycle management, matter management, e-billing, and court filing systems.
Enterprise identity integration is specified in detail, and no legal practice system is named. The agreement defines the Service as including web, API and on premise components under an order form, so a programmatic surface is contracted for. The identity layer names SAML 2.0, OIDC and SCIM 2.0 provisioning, with single sign on required on all enterprise deployments and multi factor authentication enforced through the customer's identity provider. On the isolated modes the platform runs inside the customer's own cloud estate using the customer's identity management and keys. Those are integrations with the customer's security stack, not with the systems a legal team works in. No document management system, case or matter management platform, word processor, email client or electronic signature product is named. No integrations page, connector directory or partner list is published. No API documentation is published, so the API is established from the agreement and its scope is not. Nothing describes what data moves between the platform and any system a firm already runs.
A few integrations are named among those claimed, with no catalog and no setup detail. There is no integrations page. A Plug In menu item leads to ManupatraAlert, an alert plugin on manupatrafast.com. The privacy policy says products may connect to third party services, including data visualization and cloud storage such as Dropbox and Google Drive, for uploading and sharing documents. The wider suite, including the Mykase practice management product, Manucomply and Manucontract, suggests links to the research platform. But each sells on its own domain with its own login, and nothing describes data moving between them. No practice management, document management or CRM system used by Indian firms is named. No public API, developer documentation or authentication model is published. No data flow, prerequisite or setup step is described for any connection. Nothing says whether AI Search or ManuWorks can be used from inside another application.
Deployment Model and Data Residency
Where the software runs and where the data sits. Multi tenant cloud, single tenant, private deployment, on premises, and whether region of residence is a published option or an enterprise conversation.
Tenancy and residency are both published for each mode, with what changes between them. Mode A is multi tenant software as a service hosted in India on ISO 27001 certified infrastructure, with matter level isolation. Mode B is single tenant on the L&T Vyoma sovereign Indian cloud, with private cloud isolation. Mode C runs in the customer's own AWS, Azure or Google Cloud estate, in a region it chooses, with its own identity management and keys. Mode D is on premises, from a desk unit to a full data center deployment. Mode E is air gapped, and the vendor states there is no inbound traffic at all on modes D and E. The privacy policy states data location for each mode, and residency can be locked to a chosen region on modes B to E. Customer managed keys with hardware security module storage are available on modes C, D and E. Transfers for customers outside India are governed by standard contractual clauses. The entry SaaS mode is fixed in India with no residency choice, and no subprocessor list is published to test the mode statements against.
Personal data stays in the user's country under a Manupatra commitment, but tenancy goes undescribed. The privacy policy's Data Localization section says all personal information is stored and processed on servers in the user's country of residence. Where a jurisdiction requires localization, data stays within its borders and is not moved abroad except as the law allows, with safeguards. The About page calls Manupatra a cloud first platform, and the AI brochure says the models are trained on Indian data and built in India. There are limits. The localization promise covers personal information, but Subscriber Content uploaded to the AI tools is governed by confidentiality terms that do not say where it is processed. No tenancy model is described, so nothing says whether uploaded documents sit in a shared or isolated environment. No data center, region or cloud provider is named, no on premises or private option is offered or refused, and nothing separates storage from processing for the AI features.
Security Certifications and Trust Center
Independent attestation a buyer can pull without a sales call: SOC 2, ISO 27001, penetration test summaries, a trust center with current reports and named scope rather than a badge image.
A real trust center with a named document inventory, and the reports behind a request whose terms are not stated. Ungated, the vendor's trust center shows six compliance positions: ISO/IEC 27001, ISO/IEC 20000-1:2018, SOC 2 Type 2, SOC 2, GDPR and CCPA. The vendor's own trust page adds DPDP and states the ISO certification as the 2022 revision. Audits are by CERT-In empanelled assessors, the regime recognized by the Indian government. A risk profile gives a four hour recovery time objective, and a disclosure program publishes response times of 24 hours to acknowledge and 72 to triage. Behind the access flow sit the ISO/IEC 27001 certificate, the SOC 2 Type 2 report, vulnerability assessment and penetration test reports, a security FAQ and a document titled AI Flow for Ask. The portal offers to request access but does not say whether access is instant, needs a nondisclosure click through, or follows a sales conversation. No auditor, audit period or scope statement is named. The on site trust page has five frameworks, including DPDP and omitting ISO 20000-1, while the trust center has six, including ISO 20000-1 and omitting DPDP.
No security certification or trust center appears for Manupatra. The navigation and footer have no security, trust or compliance page and no certification badge. The footer's Company links are About, Contact, Careers and Legal, and the Legal page's six tabs are Corporate, Copyright, Privacy Policy, Disclaimer, IT Compliance and Public Record. No ISO 27001, ISO 27701, SOC 2, StarAudit, CERT-In empanelment, PCI DSS or other standard is claimed, no auditor is named, and no certificate or penetration test summary is offered. Public reporting shows no certification claim by Manupatra either. This describes what is published, not Manupatra's actual security. The privacy policy lists encryption, secure servers and firewalls, access controls, and staff confidentiality agreements and training, all regularly reviewed. Clause 4.4 of the agreement commits to industry standard technical and organizational measures. Manupatra states compliance with the Information Technology Act 2000, the SPDI Rules 2011 and the Digital Personal Data Protection Act 2023, and names a Grievance Redressal Officer. A statement of legal compliance is not an independent audit.
Model Supply Chain Disclosure
Which models sit underneath, whose they are, where they run, and whether the vendor commits to telling customers when that changes. A legal buyer inherits every dependency it cannot see.
The architecture is described and nothing in it is named, and the subprocessor list the vendor says it publishes does not appear. The vendor states it built and trained its own model on its legal collection, rather than fine tuning or using retrieval augmented generation. Its model layer describes signed checkpoints, guardrails for each tenant, prompt injection sanitization and an inference audit trail. No model or model family is named. No third party model, API or inference provider is named, and nothing states whether any is used, so nothing shows whether customer content reaches an outside model on any mode. The named infrastructure is the vendor's own hosting in India, the L&T Vyoma sovereign cloud, or the customer's own AWS, Azure or Google estate, and a customer's own cloud is not a supply chain disclosure. Change notification is committed: the privacy policy says customers are notified of material changes to the subprocessor list, with a right to object for enterprise customers. But the list is stated to be published at the trust center and does not appear there.
Large language models trained on Indian data power the AI, Manupatra says, but no model or provider is named. The AI brochure and the December 2025 launch material say AI Search is powered by semantic AI and large language models. The brochure says it is trained on Indian data, built in India, and runs on a proprietary legal knowledge base developed over decades rather than generic AI trained on public data. That says what the models use and where the work was done. It does not name a model, version or provider, so nothing shows whether the models are Manupatra's own, licensed, or a third party service. Built in India describes the product, not the model. Nothing says where inference runs, no commitment to notify subscribers of model changes is published, and there is no subprocessor list. Clause 5.5 of the agreement bars sharing Subscriber Content with third parties, expressly including AI vendors and cloud providers, without written consent. That assumes such vendors exist but names none.
Commercial Transparency
Whether a buyer can learn what this costs without entering a sales process: published rates, the unit being charged, what sits behind an enterprise tier, and what implementation adds.
Published figures with a stated unit, a real free trial, and the commercial mechanics in the agreement. Two paid tiers carry figures on the vendor's site: Professional from 9,000 rupees per user per month and Enterprise from 17,250. A seven day free trial of Legal AID is published and confirmed at clause 4, requiring no payment instrument, with the vendor reserving the right to limit trial features and usage caps. Clause 2 restricts paid plans to legal or compliance professionals or people in a role requiring legal research and drafting tools. Fees, billing frequency and currency are set in an order form or product plan, all fees exclude GST, VAT and equivalent taxes, and late payment carries interest at 1.5 percent per month or the statutory rate if lower. Clause 9 publishes a 99.5 percent monthly uptime target on the SaaS mode, and clause 13 covers termination and customer content on exit. Three gaps remain. The four other deployment modes and MIRA carry no published price, no usage quota, seat minimum or overage term is published for either paid tier, and trial usage caps are discretionary.
Manupatra publishes AI prices, units and terms, so a buyer can plan costs without a sales call. The AI Solutions page has an open price table. Unlimited AI Summary costs 5,000 rupees a year for Supreme Court judgments, 5,000 rupees for High Court judgments, or 10,000 rupees for both, before GST. AI wallet plans for AI Search, AI Compare and AI Summary are 1,000 credits for 6,000 rupees, 1,500 for 9,000 and 2,000 for 12,000. The ManuWorks toolkit publishes credit costs per action: 15 a page for OCR, 3 a page for summaries, 20 a page for translation, 5 a query for document questions, 8 a page for timelines, 15 a document for comparison and 50 a query for drafting. One credit is worth 1.25 rupees. The first purchase must be at least 10,000 rupees and each later purchase at least 5,000. Activation is prorated, and a dashboard shows credits bought, used and left. The agreement sets the terms: payment in advance, no refunds, renewal and deactivation rules, price changes only from the next renewal, seven days to dispute transaction charges, and taxes on top. Two limits. The research subscription itself is not priced publicly, and the agreement says its price card is available on request. Clause 3.9 also implies a usage metered tier whose rates are not published.
Firm and Practice Coverage
Who the product is actually built for. AmLaw, midlaw, small firm and solo, in house departments, government and courts, and which practice areas are supported rather than merely claimed.
Practice scope is named specifically, and the underlying collection is not inventoried. Four practice domains are named: direct tax, indirect tax, corporate law and general laws, over a closed database of laws, circulars and judgments that the vendor calls India's largest repository of legal documents. Buyers named include general counsel, law firms and government bodies in the launch material, litigators and transactional and advisory counsel on the practitioner pages, and compliance professionals through the eligibility clause. Capability is quantified at 215 production skills that can be combined into workflows. Currency of Indian law is addressed once in substance: the vendor distinguishes the Bharatiya Nyaya Sanhita from the Indian Penal Code as a failure of general purpose assistants, a real point in a jurisdiction that recently recodified its criminal law. No court or tribunal is named, no date range is given, nothing distinguishes full text from headnotes, and nothing states which parts of Indian law are thin or out of scope. The United States preview beta announced in March 2026 has no coverage description.
Coverage is set out by jurisdiction, content and buyer, with no limits stated by Manupatra. It covers Indian case law from the Supreme Court, all High Courts and all tribunals. It adds statutes, regulations, notifications, reports, commentary and secondary sources across Indian jurisdictions and Commonwealth countries. It has a coverage page, reports more than 10 million documents, and adds new material daily. A separate Bangladesh database, BDlex, is sold alongside. The AI pricing page names students, lawyers, law firms and corporate legal teams as buyers. The About page lists practitioners, in house legal teams at companies and banks, academic institutions, government and regulators, and students and researchers across Asia. As a general case law database, it is not packaged by practice area. No limit is stated. Nothing names an out of scope court, tribunal or content type, the earliest coverage date or any gaps, or which Commonwealth countries are included. Nothing says whether the AI features cover the whole corpus, though AI Search is described in terms of judgments only.
The 12 legal signals, side by side
Recorded rather than graded. These are the questions a practitioner has to answer before a tool touches a client matter, and the answers are taken from public material only.
Client Data in Training
Can material a lawyer puts into this product be used to train a model?
The prohibition is in the agreement, and its qualifier is explained. Clause 5 of the terms of service v2.0, effective 1 January 2026 and last updated 19 April 2026, says the customer keeps all rights in documents, queries and other content submitted. It grants only a limited license to process that content to provide the Service, and states that Customer Content is not used to train shared models. The privacy policy says the same twice, for content data and for personal data.
The trust page states that these data handling commitments are contractual rather than aspirational. The qualifier is the word shared, and the vendor addresses what sits outside it: fine tuning on customer data, if offered, happens only under a signed data processing addendum and only inside the customer's own deployment boundary. That is a bounded exception with two conditions. Content retention can be reduced to zero on enterprise and the isolated modes, and on the on premises and air gapped modes the content never leaves the customer's estate.
Training on subscriber data is barred by default and allowed only with the subscriber's express written consent. Clause 5.2 of the agreement, headed Absolute Prohibition on AI Training and Internal Use, bars using Subscriber Content in any form for training, fine tuning or improving any AI or machine learning model. That covers whole or partial, anonymized, pseudonymized, aggregated or derived content. It also bars use for building algorithms, features or products, internal analytics, database enrichment, quality work, troubleshooting, research, teaching, demos or testing.
Anonymized and aggregated use, a common exception, is closed here. Clause 5.7 requires prior express written consent in a standalone written agreement for any training or improvement use, and says silence or continued use is not consent. Clause 14.3.1 repeats the ban, and clause 5.9 makes a breach material, with indemnity and injunctive relief. The agreement does not promise never to train. It promises not to train without a separate signed permission the subscriber may refuse at its sole discretion. The privacy policy states the same rule in weaker terms for user data generally.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
The customer sets the retention window and zero is reachable, with the plan level at which it becomes reachable stated. The privacy policy gives three kinds of data their own rules. Content data, meaning documents uploaded, queries submitted and outputs generated, has its retention set by the customer, defaulting to the life of the account plus ninety days for export. Usage data, meaning query logs, session identifiers, device metadata and telemetry, is kept for up to 24 months by default and is configurable on enterprise.
Account data is held for the relationship plus seven years, which the vendor attributes to record keeping rules under the Companies Act and Income Tax Act. The trust page states that retention windows can be set to zero on enterprise and on the on premises and air gapped modes. On the entry subscription the default window applies. All customer data can be exported at any time in standard formats, and on termination it is returned or destroyed at the customer's election with attestation, which clause 13 of the terms carries into the agreement.
Manupatra's agreement sets a fixed maximum: 24 hours. Clause 4.7, headed No Retention / Mandatory Deletion, requires permanent deletion of all uploaded Confidential Information, including copies, backups, AI memory, logs, caches and traces. Deletion must happen right after the request is met and no later than 24 hours after use. Written certification of deletion is available on request. Naming AI memory, logs and caches matters, because those are where data usually survives a deletion promise.
Clause 5.3 bars storing or caching Subscriber Content beyond what is strictly needed to produce the output, and all storage must be temporary, encrypted and deleted right after use. Clause 5.6 treats AI outputs containing or derived from Subscriber Content as the subscriber's confidential information, not to be stored, reused or shared. Two limits. The 24 hour rule covers uploaded Subscriber Content, not the text of a research search or the account and usage data the privacy policy describes collecting.
And the privacy policy's own retention section only promises to keep personal information as long as necessary, and refers readers to the data protection manager for periods. Where the two differ on subscriber material, the agreement governs.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
The vendor maintains and documents its own segregation model, stated at matter level rather than only at account level. The trust page lists matter level isolation on multi tenant SaaS as one of five data controls, alongside encryption, customer managed keys and hardware security module key storage. The privacy policy states that content data is accessible only to the account it belongs to. The identity layer adds what a firm needs to operate a wall: custom roles, single sign on required on enterprise, multi factor authentication through the customer's identity provider, SCIM 2.0 provisioning, and session recording on enterprise.
The isolated deployment modes remove shared tenancy altogether. How the customer administers the isolation is not published. Nothing states whether a matter boundary is created automatically or configured, whether a user can be screened from a specific matter, how a wall is applied to an existing matter, or what an administrator sees. Nothing inherits from a document or case management system, because none is integrated. A firm would have to keep the vendor's model aligned with the walls it maintains elsewhere.
Manupatra commits to access control in its contract but describes no separation model. What is documented is access on Manupatra's side. Clause 4.2(d) limits access to staff who need it and are bound by equal confidentiality duties. Clause 5.8, headed Monitoring / Human Review, bars staff from viewing Subscriber Content or AI outputs except to fix a technical issue the subscriber raised, with access logged and kept to a minimum.
User access rules exist too. Clause 11.10 requires IP based and remote access subscribers to list named Authorized Users before the subscription starts. Clause 3.5 makes the subscriber liable for all actions under its login and allows termination for password sharing. Clause 3.6 makes an organization responsible for its users, and clause 2.3 limits display to one person at a time on plans not based on IP. What is not addressed is separation between matters, clients or users within a firm.
No permission model, role model or ethical wall is described, and nothing says whether documents one user uploads to ManuWorks can be reached by a colleague on the same subscription. A shared research database has no matters, but the AI toolkit takes uploaded case documents, and that is where a wall would matter.
Third Party Request and Subpoena Notice
If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?
No public material addresses what happens when a third party demands customer data. The terms of service run to fifteen clauses, and the mutual confidentiality clause at clause 10 has no exception for compelled disclosure and no notice obligation. The privacy policy names legal obligations as a lawful basis for processing under DPDP and GDPR and describes sharing with subprocessors, but has no section on law enforcement or government requests and no notice commitment.
The trust center's open page lists a data privacy group with cookies, data breach notifications and data into system items, and no compelled disclosure item. No transparency report or law enforcement guidelines page exists. Nothing is established either way: neither a commitment to tell the customer nor a statement that disclosure would occur without notice. Two adjacent facts bear on the practical exposure. On the on premises and air gapped modes the vendor does not hold the content, and a data processing addendum is named in clause 15 as part of the agreement but is not published.
Data may be disclosed when compelled, Manupatra says, with nothing about telling the customer. The privacy policy says personal data may be disclosed where required by law, court order or government regulation, in a merger, acquisition or similar deal, or to protect rights, property or safety. Nothing says the subscriber would be told of a demand, given a chance to object, or told afterward. There is no transparency report or law enforcement guidelines page.
The two documents conflict. The customer agreement is more protective and has no exception for compelled disclosure. Clause 4.3 says Manupatra will not disclose Confidential Information to any third party without the subscriber's prior express written consent, and clause 5.5 says the same for Subscriber Content. Read literally, that allows no exception for a court order, which cannot be the intent, and no clause reconciles it.
The agreement governs where the two conflict, but its silence on legal process is not a promise of notice. So Manupatra has said what it may disclose under compulsion, in its weaker document, and nothing about telling the customer.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
The source material is identified by type, and the rights basis is asserted only as to the vendor's own holding, with no update cadence published. It is a closed database of laws, circulars and judgments, which the vendor calls India's largest repository of legal documents, covering direct tax, indirect tax, corporate law and general laws. The vendor distinguishes it from retrieval on the open web, and traces the collection to decades of legal documentation and knowledge structuring work by named individuals in its founding lineage.
Clause 6 of the terms states that the collection, with the software, models, skill library and metadata schemas, is owned by or licensed to the vendor. That asserts the vendor's own title and does not state the basis on which the underlying legal materials are used. Nothing identifies the reporters, publishers or official sources of the judgments and circulars, or whether Indian primary law is treated as government material outside copyright or licensed from a publisher.
No update cadence is published, so nothing shows how quickly a new judgment, circular or amendment arrives. No coverage inventory exists to test the size claim.
Sources are identified, and the agreement states the licensing basis for third party material. Clause 10.2 says some information belongs to content providers, is marked as such, and is licensed to Manupatra to store, catalog and distribute to subscribers, with no rights passing to the user. The Public Record section of the legal page says public records come from courts and government agencies. Manupatra collects them from established government and judicial sources, in some cases under agreements whose terms may be stricter than the law.
The copyright section says third party content is reproduced with prior permission and its copyright retained. The AI brochure grounds the models in Manupatra's own edited corpus, built over decades from citations, judicial treatment and summaries, which it owns. So a buyer can see the types of source and how each is held. Not published: a list of suppliers, any license term or duration, or whether the licenses cover AI processing as well as distribution. That matters for a licensed corpus feeding a language model.
Good Law Verification
Does the product tell you when the authority it just cited has been overruled?
No public material addresses whether authority is checked for later history, and the verification the product does perform answers a different question. The product verifies existence and support. Every response lists the sources relied on, each openable and downloadable with a summary and a statement of why it was pertinent, and the reasoning chain flags a claim whose trace to a source is broken. That shows that a cited judgment is real and says what the vendor claims, not that the judgment still stands.
Nothing states whether a decision has been overruled, reversed on appeal, stayed, distinguished, declared per incuriam or superseded by statute. No citator is licensed, no treatment signal is computed or shown, and no currency flag attaches to any authority. The vendor does contrast general purpose assistants that treat the Bharatiya Nyaya Sanhita as though it were the Indian Penal Code, a statutory currency point, and no mechanism is described for either kind of currency.
Treatment flags come from a system of Manupatra's own rather than a licensed one. The product page section titled Be sure you are using Good Law says effective date highlights and blue, yellow and red flags mark whether each statute and judgment is good law. That flag scheme, on both statutes and judgments, works as a citator and is Manupatra's own. Nothing suggests any part is licensed, and the AI brochure describes the treatment data as part of the edited corpus built over decades.
The AI uses it. The brochure says AI Search lets a user see whether a judgment has been overruled, distinguished or followed, find conflicting judgments, and trace how courts have read a statute over time. The product page lists authority checks and citation maps among its analytics. Two limits. No method says how treatment is decided, whether by the advocates who edit the corpus or by software, and nothing says which courts, years or content types carry flags. Nothing says what a flag means in borderline cases or how fast a new adverse judgment updates the flag.
Refusal and Uncertainty Behavior
What does the product do when the answer is not in the corpus?
What the system does when it cannot support a claim is stated in terms, without a demonstration behind it. The vendor commits that every factual claim in every output is traced back to its source and that a claim whose trace is broken is flagged, with no exceptions. That describes behavior on failure. Two further mechanisms sit alongside it. The model layer names output hallucination detection and a meta reasoning gate through which every input and output passes, and the vendor states that failure of any layer opens a ticket rather than a breach.
MIRA is described as verifying its own reasoning against a layer built to catch its mistakes before it produces output. No evaluation, benchmark, worked example or published result shows the flagging operating, and no rate is given for how often a trace breaks. Nothing states what the user sees when a claim is flagged, whether it is withheld, marked or returned with a caveat. A trust center item is titled A note on Hallucinations.
When the system cannot give a reliable answer, nothing says what it does, though Manupatra gives useful guidance on where the tool is weak. Its AI brochure section titled When NOT to Use AI Search tells users to avoid it for exact queries, party names, citations, exact statutory wording and verbatim quotes. It says AI Search does not follow Boolean logic, and routes those queries to MANU SEARCH, LEGAL SEARCH and CITATION SEARCH.
Another page shows how to frame a good query, with good and poor examples. All of that is about what the user should do before a query, not what the system does after one. Nothing says AI Search declines a query, flags low confidence, marks an uncertain result for checking, returns fewer than ten results when fewer are relevant, or says it found nothing. It always returns a ranked ten, so whether the tenth result is relevant or merely nearest is unanswered.
Fabricated Citation Record
Does a public court record exist addressing fabricated or hallucinated legal citations in output from this product?
The AI Hallucination Cases database maintained by Damien Charlotin, and Indian and international reporting on sanctions and hallucinated authority, record no court order, opinion or disciplinary record naming Lexlegis.ai or Lexlegis Solutions Private Limited. The Indian decisions on the record name no tool. Indian courts are actively engaged with the question. In a February 2026 order, the Supreme Court took cognizance of a trial court that relied on nonexistent judgments generated by AI, describing it as a matter of institutional concern bearing directly on the integrity of the adjudicatory process and indicating that such a decision would be misconduct with legal consequences to follow.
A separate bench has raised petitions drafted with AI that cite nonexistent authority. The vendor sells into that market on an express claim of being hallucination free, and nothing on the public record to date names it.
The AI Hallucination Cases database kept by Damien Charlotin holds no entry naming Manupatra or Manupatra AI Search as the tool used, and no Indian decision on fabricated citations produced with AI names it as the source, including the Supreme Court of India's July 2026 ruling. In the commentary on those decisions, Manupatra is named often as one of the authorized databases for checking citations.
Bar Guidance Alignment
Has the vendor engaged in public with the ethics opinions its buyers are bound by?
Professional responsibility is engaged squarely in the agreement, and no bar or court instrument is named. Clause 8, headed AI outputs and professional responsibility, says outputs support legal or compliance work rather than giving legal advice. The customer must review and validate all outputs before relying on, filing or acting on them, no output creates a lawyer client relationship with the vendor, and professional judgment is not replaced.
Clause 7 prohibits using outputs to falsely attribute authority or to deceive any court or tribunal, which addresses the conduct that professional guidance on AI is written to prevent. Clause 2 limits paid plans to legal or compliance professionals and people in roles requiring legal research and drafting tools. No Bar Council of India rule, provision of the Advocates Act 1961, state bar council circular, or court practice direction or standing order on AI assisted filings is cited, mapped or linked. Indian courts have publicly begun addressing AI generated authority.
Clause 16.1 of the agreement, headed No Legal Advice, says material on the site is not legal advice and creates no relationship of client and advocate. The legal page's disclaimer repeats this for forms. Clause 11.12 requires judgment text to be verified against the certified copy, and clause 15.3 makes the user responsible for checks. No bar body, conduct rule or ethics opinion is cited, Bar Council of India guidance is not mentioned, and nothing maps any AI feature to a professional duty.
In July 2026 the Supreme Court of India held that citing unverified AI generated precedent is professional misconduct. It directed India's Bar Council to form a committee to frame norms, with draft rules on AI in courts out for consultation. Manupatra publishes that ruling as news on its own legal updates service. It says nothing linking AI Search, AI Summary or the AI Drafter to the verification duty that ruling places on each lawyer.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
Time savings are marketed to a buyer who bills clients, and nothing addresses the client's side of the bill. Law firms and litigators are named buyers alongside general counsel and government bodies, and the work being compressed is legal research and drafting, which a firm bills. The savings claims are explicit: research memos in hours rather than days, hearing preparation without the all nighter, authority bundles built automatically, and answers in seconds.
No published material addresses billing, fee treatment or client disclosure. Nothing tells a firm how to treat research hours that collapse, no record of AI assisted work for each matter is offered as a basis for a fee narrative, and no guidance on disclosing AI assistance to a client is published. Clause 3 of the terms governs only what the customer pays the vendor and never reaches the customer's own invoicing. The vendor engages the lawyer's other professional obligations at length, gating eligibility to professionals and prohibiting the use of outputs to deceive a court.
What happens to the bill when AI cuts the time a task takes goes unaddressed by Manupatra. The product sits inside the fee relationship. Its pricing page names lawyers, law firms and corporate legal teams as buyers, and in Indian practice research is usually billed as time or passed on as a cost. Nothing says which applies. No per matter record of AI work is described, nothing marks output as AI assisted for billing, no guidance on fees or disclosure is published, and no saving is claimed in billable terms.
The AI brochure's benefits are about finding what keyword search misses, not time saved. The credit wallet is very detailed about cost, with per page and per query rates, a fixed credit value and a dashboard of credits bought, used and left. That could let a firm assign AI cost to a matter, and clause 3.9 gives a detailed statement of transaction charges. But that is technology cost, not AI assisted work, and Manupatra never presents it as a billing or disclosure tool.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
The documents a firm would forward are named and governed, and the one said to be published does not appear. Clause 7 of the privacy policy says the vendor relies on a small set of subprocessors, that a current list is published at the trust center, that customers are notified of material changes, and that enterprise customers may object. Clause 15 of the terms names a Data Processing Addendum and a Security Schedule as part of the entire agreement, so both exist and are incorporated.
The trust center carries a documents flow through which they and the certification reports can be requested. The subprocessor list does not appear on the open view of the trust center, so the commitment to notify changes attaches to a document a prospective customer cannot read. No model provider is named, so a firm cannot tell its client whose model, if any beyond the vendor's own, processes matter content. The Data Processing Addendum and Security Schedule are named and not published, which puts the forwardable material behind a request.
None of the three items a firm would forward is available from Manupatra: no subprocessor list, no named model provider, and no client facing consent or notice pack. The AI is described only as semantic AI and large language models trained on Indian data and built in India. A firm asked who processes its uploaded documents could not answer from a list. No data processing addendum is published separately, and nothing is offered on request.
What a firm does have is a published customer agreement it can forward, which answers the core question firmly. Clause 5.5 bars Manupatra from giving access to, disclosing or transferring Subscriber Content to any third party without express written consent. That includes affiliates, subcontractors, service providers, AI vendors, cloud providers and other users. Clause 4.3 says the same for Confidential Information, clause 5.2 bars training, and clause 4.7 requires deletion within 24 hours.
On the question a client's AI clause asks, that is a strong answer. It is still not a subprocessor list or provider statement, which is what a firm's guidelines usually ask for.
Court Disclosure Support
If a judge’s standing order requires an AI disclosure, can the product produce one?
Real elements of a record exist, and none is offered as something a lawyer could produce. The strongest is specific to AI work: the model layer states an inference audit trail in which every call is logged and attributable. That records what the model did and who invoked it, not who opened a file. Every output also carries the sources it relied on, openable and downloadable with a statement of why each was pertinent, and the reasoning chain flags any claim whose trace to a source is broken.
Clause 7 of the terms prohibits using outputs to falsely attribute authority or deceive a tribunal, which shows the scenario was considered. The disclosure half is absent. Nothing states that the inference audit trail is exportable, available to the customer rather than the vendor, retained for any period, or presentable in a form a court would accept. No certification, template, declaration or guidance on a practice direction about AI assisted filings is published.
Manupatra publishes nothing that helps a lawyer disclose AI involvement to a court, and the gap matters here. Two related features exist, and neither is a disclosure tool. The AI brochure says the right panel gives an AI generated analysis of the query that can be downloaded or shared, and results can be emailed, printed or downloaded. But nothing says an exported analysis is labeled as machine generated, carries a timestamp, records the query, or can be told apart from a manual note once it leaves the platform.
Clause 11.12 requires judgment text to be checked against the certified copy, which concerns the accuracy of the source, not the model's involvement. There is no audit trail of AI use, no per matter or per query record, no certification template, no export built to show machine involvement, and no guidance on when to disclose AI help. In July 2026 the Supreme Court of India held that citing unverified AI generated precedent is professional misconduct, and that a decision resting on hallucinated material is no decision in law.
Indian advocates may soon need to show how a cited authority was found. This platform produces AI case analysis and gives no way to show later which words were the model's.
In summary
Lexlegis.ai
Lexlegis.ai, from Lexlegis Solutions Private Limited of Mumbai, is an Indian legal AI platform built on a closed corpus of Indian laws, circulars and judgments. Legal AID answers questions on direct tax, indirect tax, corporate and general law with sources, reads uploaded documents and drafts submissions. MIRA runs multi step work from 215 skills. According to the AI Legal Index, Lexlegis.ai offers five deployment modes up to air gapped, states ISO 27001 and SOC 2 Type 2, and prices from ₹9,000 per user a month. It names KPMG and the Government of India among customers. Each claim traces to its source, and broken traces are flagged.
Manupatra
Manupatra, from Manupatra Information Solutions Pvt. Ltd., founded in 2000 and based in New Delhi and Noida, is India's oldest online legal research platform. It covers the Supreme Court, all High Courts and tribunals across more than 10 million documents, with its own good law flags. Its AI Search, launched in December 2025, retrieves judgments by concept, alongside AI Summary, AI Compare and the ManuWorks toolkit. According to the AI Legal Index, Manupatra publishes prices for its AI add ons, and its agreement bars training without written consent and requires deletion within 24 hours.
Questions buyers ask
Lexlegis.ai vs Manupatra: which is better for Indian legal research?
Both put generative AI over Indian case law and statutes. Lexlegis.ai leads on deployment options, security reports and named customers. Manupatra brings a 25 year old corpus with its own good law flags and stricter contract terms on uploaded material. Researchers who need to check whether a judgment still stands have more to use in Manupatra. From the AI Legal Index, based on each vendor's own published materials as of September 27, 2026. No vendor pays for placement.
Does either tool check whether a judgment is still good law?
Manupatra does. It marks statutes and judgments with blue, yellow and red flags and effective date highlights, and its AI Search can show whether a judgment was overruled, distinguished or followed. Lexlegis.ai traces each claim to its source and flags broken traces, which confirms a source exists. From the AI Legal Index, based on each vendor's own published materials as of September 27, 2026. No vendor pays for placement.
Can Lexlegis.ai run on premises?
Yes. Lexlegis.ai publishes five deployment modes: multi tenant cloud in India, single tenant on an Indian sovereign cloud, inside the customer's own AWS, Azure or Google estate, on premises, and air gapped with no inbound traffic. Region locking and customer managed keys are available above the entry mode. Manupatra is delivered as a cloud service. From the AI Legal Index, based on each vendor's own published materials as of September 27, 2026. No vendor pays for placement.
How much do Lexlegis.ai and Manupatra cost?
Lexlegis.ai publishes Professional from ₹9,000 and Enterprise from ₹17,250 per user per month, with a seven day free trial. Manupatra publishes its AI add ons: AI Summary at ₹5,000 a year per court level, credit packs from ₹6,000 for 1,000 credits, and a per action rate card for its toolkit. Manupatra's base research subscription price is available on request. From the AI Legal Index, based on each vendor's own published materials as of September 27, 2026. No vendor pays for placement.
Neither vendor paid for inclusion, placement or a grade, and neither reviewed this page before it published. Everything on it comes from public material on the dates shown. How the index grades.
Three points to weigh. Both vendors claim their AI is free of hallucinations, and neither publishes an accuracy or error measurement. Lexlegis.ai's commitment is not to train shared models, with fine tuning on customer data possible under a signed addendum inside the customer's own deployment. Its policy points to a subprocessor list that is not at the location the policy gives. Manupatra claims no security certification, and its privacy policy permits disclosure under legal compulsion without customer notice. Neither vendor reviewed this page.