Lexzur vs Xakia: how they compare in 2026
Lexzur and Xakia overlap in legal matter management, but the surrounding workflows lead to different choices. Xakia is built for internal legal departments organizing requests, tasks, budgets and outside counsel spend. Lexzur also serves law firms through PRACTICE, combining matters with client billing and accounting, while CONTRA supplies a separate contract lifecycle product. Lexzur's LEXA supports configured agents and linked workflows across that environment. Xakia applies AI to functions including contract review and invoice analysis within its department toolkit. A legal department seeking matter visibility and control of incoming legal bills has a direct use for Xakia. An organization that also needs firm financial administration, Arabic work or customer server deployment has a closer match in Lexzur. Their user rates should be compared alongside AI allowances, portals and implementation scope rather than as identical bundles.
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.
LEXA operates within CONTRA and PRACTICE, bringing contract analysis, matter summaries, translation and drafting into the same records used for daily legal administration. Teams define agents with persistent instructions, reference files and tools, then combine them into sequential, dependent or routing workflows. A workflow can call another workflow and be enabled or disabled centrally. The Knowledge Base supplies organizational templates and policies, while chat settings select jurisdiction and response preferences. The underlying products still manage matters, agreements, permissions, billing and signatures without generating answers. AI is consequently an important working capability within a broader operational system, rather than the sole function of the platform.
Xakia's AI powers contract review, term extraction, summaries, plain language retrieval and invoice review within a broader legal operations platform. Contract reviewers apply a chosen playbook in Word and select Pragmatic, Balanced or Strict review settings. Extracted terms retain the source clause and require approval before entering the repository. Invoice analysis applies billing guidelines and flags inconsistencies in totals, tax and vendor details. AI remains optional, with extra charges on Advance and Professional and inclusion on Enterprise and All In. The underlying matters, tasks, intake, spend and entity workflows can operate without models. Connections from Claude and Copilot remain forthcoming, initially with read access only.
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.
LEXA compares agreement language with the customer's templates, clause library and policies, presenting the clause or passage behind each finding. Review covers missing provisions, unusual wording, indemnities, obligations and ambiguity. Administrators can upload reference files, index selected material and apply custom instructions across responses. This gives analysis an identifiable organizational context instead of relying only on a general question. Matter checks also use the file's correspondence and attachments. No measured citation accuracy, described legal evaluation set or error rate accompanies these features. A retrievable contract passage supports inspection of a finding but does not verify an external legal authority or its continuing validity.
Xakia's AI points back to the customer's own documents. Key terms extraction returns each term with a confidence indication, the clause it came from and the exact text it relied on. A user reviews each value, corrects it and clicks Apply before anything reaches the contract repository. Contract review marks up a third party contract in Word against the customer's chosen playbook, governing law, market practice and basic drafting standards, at a Pragmatic, Balanced or Strict setting. It flags each deviation with a suggested redline for the lawyer to review. The query generated from a plain English request remains visible, editable and reusable. The AI addendum says output may not be accurate or up to date and must be reviewed by a lawyer. The FAQ commits to expanding AI without compromising accuracy, and Xakia gives no accuracy figure, error rate or test results.
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.
Administrators configure each agent's task, response behavior, available tools and reference files. Workflows run sequentially, pass earlier results to later agents or route work to an appropriate agent or subworkflow; each can be disabled from settings. Users select the workflow and supporting material, with an optional buyer, seller or neutral review perspective. LEXA returns source passages for review and records AI actions against the underlying record. Its stated boundary excludes signing, approving and making the final legal risk decision. The agreement requires human verification and approval before reliance, execution or filing, and permits binding AI action only through a separate written arrangement.
Every Xakia AI feature requires a human decision before material is saved, applied or sent. Contract review proposes revisions inside Word against the selected playbook, and a reviewer decides which changes to use. Extraction exposes each value and its supporting clause before the user presses Apply. Invoice reviewers can waive findings, and their decisions enter an audit trail. The AI addendum requires legal practitioner supervision and prohibits consequential automated decisions without proper human review. These are usable review mechanisms, but the service does not specify a general stopping rule for unsupported reasoning or a complete correction and escalation process after an erroneous result.
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.
Al Marbae & Partners Law Firm in Jeddah began using PRACTICE in December 2023. Its administration and business development manager, Mohammad Ali, describes improved efficiency, reduced operating effort and more accurate billing. The account identifies the firm, responsible user, product and adoption period, but supplies no numerical baseline or measurement method. CONTRA separately promotes faster contract creation, lower costs and extensive approval automation without connecting those figures to that firm's experience. The named deployment therefore supports actual operational use of Lexzur, while the size of any improvement attributable specifically to LEXA remains undetermined. General customer totals do not establish AI performance.
Coca-Cola Europacific Partners uses Xakia to consolidate the Australian legal team's matters and replace paper forms and monthly spreadsheet reporting. Its deputy general counsel describes reporting delays of up to two months before adoption and current visibility into matters afterward. The team also used the information to identify trademark workload and allocate that work to an internal specialist. Other named users describe workload allocation, request handling and integrated files, including Contact Energy, Kiwi Property and the City of Murfreesboro. These are operational accounts from identifiable organizations. They do not measure the contribution of current AI features or provide a controlled before and after study of contract or invoice analysis.
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.
The agreement protects confidential information through purpose restrictions, reasonable care and obligations on authorized recipients. Lexzur separates customer databases and attachment paths, with privacy settings for individual matters and folders. LEXA follows the workspace's user permissions, while account administrators retain broad access. Product guidance states that customer files do not train Lexzur models or improve others' outputs. The contract nevertheless permits derived anonymous statistics for development and improvement, and the data addendum includes system improvement among AI processing purposes without an express training exception. Provider retention, a complete processor schedule and privilege specific handling remain unresolved, so the marketing commitment does not supply the entire contractual boundary.
Xakia restricts confidential matters to designated people and applies those restrictions to files in its integrated repository or connected document system. Roles, teams, locations and individual permissions further limit access, while outside counsel receive only material shared through Connect. The service agreement protects subscriber information, and the AI addendum prohibits training on inputs and outputs by Xakia or its model providers. A separate clause permits training on anonymous data and statistics derived from inputs. Inputs are processed transiently, with no Xakia retention after delivery. These controls address matter confidentiality and privilege, but unnamed third party models and their specific handling remain unresolved.
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.
LEXA is presented as assistance with review, extraction, summaries and drafting, with the lawyer retaining professional judgment. The agreement excludes legal and regulatory advice and requires independent verification and approval before generated material informs filings, contracts or other decisions. The product also serves procurement, sales and other business users, so this boundary extends beyond licensed lawyers. A country selector provides jurisdictional context, and a general framework applies when no country is selected. That setting does not establish jurisdiction specific authorization, professional competence or legal correctness. No detailed allocation of supervision duties across those countries accompanies the available workflows.
Xakia's AI output is general information rather than legal advice. Its AI addendum requires use by legal practitioners or under their direct supervision, in accordance with professional conduct obligations. Users must validate or disregard output using their own skill, judgment and experience, and a lawyer must review it as appropriate to the intended use. Automated decisions with legal or similarly significant effects require human review. These obligations address professional supervision alongside the product's drafting and analysis functions. They do not set jurisdiction specific limits on what legal work the models can perform or map the controls to particular bar guidance.
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.
The security exhibit lists grounded answers, citations, estimates of certainty, approval mechanisms and monitoring as optional AI controls, applied where appropriate. General operational provisions cover development changes, testing and security procedures without a defined AI release assessment. Customer administrators can constrain agent instructions, reference files and tools, and can disable workflows; those settings govern customer use rather than establishing the vendor's own testing program. No accountable AI governance owner, model acceptance criteria or results across legal tasks and populations are identified. Educational discussion of bias and human judgment does not supply a measurable assurance process for current LEXA agents.
Xakia's AI controls sit in the features and the contract rather than in a governance framework. Its AI page makes every feature opt in, so the customer decides what its team uses and when, and gives each feature a human checkpoint so nothing is applied without a user's sign off. The AI addendum bars automated decisions with legal or similarly significant effects on individuals without proper human review, and bars uses that are discriminatory or bias inducing. It also requires a lawyer to review output. The feature set includes reviewed contract revisions, extracted terms and invoice findings, but no evaluation method accompanies those workflows. Xakia sets out no AI principles, owner of AI governance, testing before release or bias findings, holds no ISO/IEC 42001 certification, and its security page covers information security.
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.
Lexzur combines encrypted storage and transport with user permissions, access logging and backup procedures. Its data addendum requires notice without undue delay after a confirmed personal data incident, aiming for seventy two hours where feasible. The agreement sets deletion within ninety days after termination, but the addendum preserves legal, compliance and temporary backup exceptions. Chat History retains conversations for later continuation, without a separate prompt retention window. Subprocessors must accept protective obligations, although their full list is supplied on request and change notice is conditional on applicable law. These controls cover substantial operational ground without a complete, fixed account of every AI processing layer.
Xakia's agreement retains subscriber data for thirty days after termination before deletion, while the AI addendum treats input and output processing as transient and leaves necessary retention to the customer. Backups run every five minutes and remain for thirty five days in the same country. The DPA requires breach notice without undue delay and thirty days' advance notice of new subprocessors, with objection rights. Encryption, restricted matters, multifactor authentication and annual penetration tests support access protection. The processor schedule identifies hosting and business systems but does not name the AI model providers. The service terms also exclude responsibility for sensitive data, including protected health information.
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.
The agreement provides a vendor defense and indemnity for infringement claims concerning the program and services, with exclusions for customer material, modifications, misuse and combinations with other products. It limits aggregate liability to fees paid during the preceding twelve months and excludes consequential losses. AI output carries no accuracy, completeness or legal sufficiency warranty, and the customer must verify its use. Service interruption credits are a separate, limited remedy, capped at five percent of the monthly fee under the stated availability schedule. These contractual remedies address defined service and intellectual property risks, without a distinct promise to compensate for an erroneous legal conclusion.
Xakia's platform agreement and AI addendum apply warranty disclaimers and liability limits to the AI features. Each party's liability is capped at the amounts paid or payable in the 12 months before the first incident. Xakia's liability for breaching its data security or confidentiality obligations, or any data processing addendum, rises to three times that cap. Indirect, consequential and lost profit damages are excluded. Its intellectual property indemnity is outside the general cap. Xakia defends the subscriber and its users against third party claims that the platform, used as authorized, infringes intellectual property. Trials and betas carry no warranty or indemnity, and Xakia's liability for them is capped at $1,000. The AI addendum says output may not be accurate or up to date and must be reviewed by a lawyer, and Xakia gives no indemnity specific to AI output.
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.
Lexzur webhooks send outbound HTTPS POST requests when matters, intellectual property records, matter containers or contracts are added or edited, and when supported workflow statuses change. Administrators configure a primary endpoint and optional second endpoint, then trigger an event to verify delivery. The receiving systems must process the event data. Microsoft connections associate Outlook mail with matter records and support Word work, Teams and external storage. Signature integrations return signed copies and audit trails to the contract record. Google Calendar synchronization exchanges calendar, meeting, attendee and timing information in both directions. API availability and some integrations depend on the purchased plan.
Xakia's pricing page lists its integrations by plan. Every plan includes Outlook, Word, SharePoint, Gmail and Microsoft Teams integrations, API access, and single sign on through Entra, Okta, SCIM or ADFS. Professional and above add iManage and NetDocuments integrations and connections to Brightflag, DocuSign, Ironclad and Josef. Slack and Salesforce appear on the higher plans, with Salesforce at extra cost. Xakia offers to keep documents in the customer's own SharePoint rather than migrating them, and the API carries no charge. The pages name each connection without saying what data moves in which direction or what an administrator configures. An MCP server with connectors to Claude and Copilot is listed as coming soon for All In.
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.
Lexzur offers cloud hosting, private service, hybrid arrangements and installation on customer servers. The cloud description identifies Microsoft Azure, while the data addendum names Saudi Arabia, the United Arab Emirates and the European Union among available hosting locations. Product guidance places local residency on Business and higher plans, with the chosen arrangement defined in the order or proposal. Customer databases and attachment paths are separated. LEXA uses Azure OpenAI, but its inference region and the treatment of remote support or other processors are not mapped to each hosting choice. Keeping the core repository in a selected country therefore does not establish identical residency for every AI operation.
Xakia runs on Microsoft Azure, and customers choose where data is stored in Australia, Canada, the Netherlands, the United Kingdom or the United States, with a Middle East location listed as coming soon. Different users in one account can have different data locations. European and UK customers can keep data on Azure servers in the United Kingdom or the Netherlands, and the Terms of Service, effective 1 October 2025, are built around the data location the subscriber selects. Section 1.4 of the AI addendum says AI processing runs where the third party AI tools are available and may differ from the selected location, with the processing location shown in the subscriber's account. Separate privacy policies cover Australia, the United States, Europe and Canada. Whether customers share infrastructure is not stated.
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.
Lexzur identifies ISO/IEC 27001:2022 certification with certificate number AE24060703. Its trust center lists ISO certification and separate SOC 2 Type II compliance entries for Saudi Arabia, France and the United Arab Emirates. It also lists resources covering encryption, backups, disaster recovery, secure development and security fixes. Detailed reports and control material require access; the open overview does not supply auditor opinions, examination periods or full service scope. Other explanations attribute SOC certifications to the hosting provider, which is a separate assurance boundary. Encryption, authentication and role controls supplement these statements without independently proving the quality of AI answers.
Xakia states certification to ISO/IEC 27001:2022, SOC 2 Type 2 and HIPAA compliance on its security page, FAQ and pricing page, with regular independent audits of its applications, systems and networks. Its trust center shows compliance monitoring and controls and offers the ISO 27001:2022 certificate, a penetration test attestation and the SOC 2 Type 2 report on request. The data processing agreement and current and past subprocessor lists sit beside them. Third party experts test the application and infrastructure once a year, and third party tools scan the platform for vulnerabilities continuously. The platform is hosted on Microsoft Azure in Tier III facilities certified to ISO 27001:2022, SOC 2, PCI DSS, HIPAA and FedRAMP, with Australia's IRAP and the UK's G Cloud among the regional standards. The open pages name no auditing firm, audit period or report date.
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.
LEXA identifies Microsoft Azure OpenAI as its generative AI technology. The application allows agents to use Precise or Creative response behavior, but those labels do not identify foundation model names or versions. Administrators also choose reference files and available tools, shaping context without selecting a fully described provider chain. The agreement permits third party model and infrastructure services with their own technical and regional constraints. It does not promise advance notice of particular model substitutions, and the complete subprocessor list is available on request. The provider is therefore identifiable, while model versions, processing regions and change handling remain only partly specified.
Xakia's AI addendum refers to third party AI tools, defined as the third party language models and AI technology behind its AI features, without naming any of them. It says Xakia does not control their operation or continued availability, and that their processing location may differ from the subscriber's data location and is shown in the subscriber's account. Section 6 says any flow down terms the providers require are displayed at trust.xakiatech.com, and section 8.3 gives at least thirty days' notice of changes to the addendum. The subprocessor list in Xakia's trust center names Microsoft and business systems such as HubSpot and Stripe but no AI model provider. Neither the third party AI tools nor Xakia may train models on customer inputs or outputs.
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.
Both PRACTICE and CONTRA use subscriptions priced per user, with cloud Basic starting at $9 per month billed annually or $11 monthly. Basic covers three to five users; Business supports up to twenty five and Enterprise up to fifty, with Enterprise Plus quoted for larger arrangements. Annual Business and Enterprise rates are $29 and $39 per user per month. LEXA includes fifty trial credits per instance, followed by purchased transaction bundles. Portals and selected connectors add separate charges. Implementation scope and residency require a proposal. The subscription and usage structure is clear, but the full cost depends on deployment, external access and the purchased services.
The interactive pricing schedule lists Advance at $100, Professional at $125, Enterprise at $155 and All In at $230 per user monthly, with USD as the default currency and alternatives for other regions. Advance adds $50 per user monthly for AI and Professional adds $30, with usage caps. Enterprise includes capped AI and All In removes caps subject to fair use. The linked machine readable schedule still gives lower base prices and older availability descriptions. Onboarding has a separately quoted fee. Monthly service periods require thirty days' nonrenewal notice under the agreement, while annual billing offers a discount. Outside counsel use Connect without a portal or invoice percentage charge.
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 serves law firms and corporate legal departments handling matters, litigation, hearings, contacts, time, billing and intellectual property administration. CONTRA extends agreement workflows to procurement, sales, finance, human resources and government teams. Plan sizes range from three users to enterprise arrangements, with separate portals for clients and outside counsel. The interface includes English, Arabic, French and Spanish, and LEXA supports Arabic review and translation. Regional features include Saudi electronic invoicing and local signature options. A jurisdiction setting gives AI additional context, but no complete matrix identifies supported legal authorities, practice specific performance or excluded uses across all offered countries and languages.
Xakia targets in house legal teams of roughly 2 to 200 people, a range it repeats on its pricing page. The platform can be configured in English, Japanese, Spanish, French, Danish, Dutch and Brazilian Portuguese, and the spend module supports all currencies. Xakia has offices in Melbourne, Kansas City and London and says it has many hundreds of clients worldwide. Scope is stated by module, covering matters, intake, contracts, spend, documents, entities, disputes and reporting, with a stakeholder license for non legal users. Customers store data in Australia, Canada, the Netherlands, the United Kingdom or the United States and can set different data locations for different users. Industry segments and practice areas are not described.
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 October agreement permits aggregated, anonymized statistics for research, system improvement and product development. The data addendum also allows AI processing of customer data for system improvement and functionality enhancement, without an express exclusion for model training. This is broader than LEXA's promise that customer files do not train models or improve others' answers. The Knowledge Base uses a Train action to index reference files; that interface label alone does not establish foundation model training. The contractual improvement permission has no stated opt out.
Xakia's AI addendum bars training on customer content and permits training on anonymized data derived from it. The addendum, effective 1 February 2026, forms part of the Terms of Service. Section 2.4 says neither Xakia nor its third party AI tools will train any AI model on inputs or outputs, and that using them to improve service features is not training. Section 2.5 lets Xakia collect derived data, defined as anonymous data and statistics derived from inputs.
Xakia may use that data for benchmarking, reports and training its AI features or AI models, and discloses it outside Xakia only in anonymized, aggregated form. Section 1.3 lets a customer opt out of the AI features in its account, and the addendum offers no separate opt out from derived data. Xakia does not name the third party AI tools behind the features, and their processing may run outside the customer's selected data location, as shown in the customer's account.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
LEXA Chat History lets users reopen and continue past conversations, confirming that the application retains exchanges. No distinct prompt or output retention period, configurable window or zero retention option is specified. The agreement's ninety day deletion period follows subscription termination rather than each AI request. Its data addendum permits legal and internal compliance retention and temporary copies in backup or recovery systems.
Those exceptions mean the termination period cannot be treated as an absolute purge deadline for all conversation material or provider records.
Section 2.8 of Xakia's AI addendum sets the retention period for AI prompts and outputs at zero. Because the AI features process data ephemerally, Xakia does not retain inputs or outputs after the output is delivered. Inputs are processed transiently and not stored, and the customer is responsible for capturing anything it needs to keep. Xakia fixes the period rather than letting the customer configure it. It governs the AI features rather than matter data held in the platform, which the Terms of Service keep for 30 days after termination before permanent deletion.
Extracted key terms enter the contract repository only when a user clicks Apply, and then follow the matter data's retention.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
Lexzur controls access through user groups and module permissions, with additional privacy at matter and folder level. Administrators can assign permissions for individual activities in PRACTICE, CONTRA and Billing; administrators have all permissions by default. Client and advisor portals provide separate, restricted participation, including field controls for outside counsel. LEXA follows the workspace's permissions.
These are Lexzur's own controls rather than demonstrated inheritance of an external document system's ethical walls. Knowledge Base administration also requires care because indexed reference material can influence responses across modules.
Role based permissions control access by user, team or matter, Xakia says sensitive matters can be restricted to specific people, and outside counsel see only what the customer chooses to share through the Connect portal. Single sign on through the customer's own identity provider handles authentication. The product runs its own permission structure rather than applying a document management system's access model at query time, and Xakia does not say whether the AI features apply those permissions when they read documents.
Third Party Request and Subpoena Notice
If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?
The confidentiality clause requires reasonable advance notice of compelled disclosure where legally permitted and cooperation with protective measures at the customer's expense. For personal data, the data addendum promises commercially reasonable efforts to notify before disclosure where lawful and reasonably practicable. Authorities may still require production, and the commitment is subject to those legal and practical qualifications.
No transparency report describing request volumes or handling outcomes accompanies the terms. Incident notification addresses unauthorized access separately and should not be confused with notice of legal process.
Xakia's data processing addendum commits it to inform the subscriber promptly of any third party request made directly to Xakia about its processing of the subscriber's personal data, with details of the request. The commitment applies to the extent the law permits. Third party requests include those from a regulatory authority. Xakia will not respond to such a request without prior notice to the subscriber and a chance to object, except where the law requires it or to confirm that the request relates to that subscriber. The commitment covers personal data under the addendum, and Xakia issues no transparency report.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
LEXA lets a user select a country whose legal framework should guide the response, with a general framework used when none is chosen. It also analyzes uploaded agreements, matter files and organizational reference material. The country setting does not identify the underlying statutes, judgments or commentary, their licensing basis or their update schedule. Customer templates and policies provide a distinct source of organizational context, but they are not a complete primary law collection. No inventory maps the available jurisdiction settings to named, maintained legal corpora.
Xakia holds no legal content collection. Its AI features work on the customer's own contracts, documents and invoices in the platform, so the material and its provenance are the customer's. Contract review applies the customer's own playbook and also checks a contract against governing law, market practice and basic drafting standards, without naming where that knowledge comes from. Xakia supplies no licensed legal content and states no license basis or update cadence for the knowledge its model applies.
Good Law Verification
Does the product tell you when the authority it just cited has been overruled?
LEXA can analyze legal issues, governing law and compliance questions using matter material and selected jurisdictional context. Contract findings point back to relevant clauses, supporting review of the working agreement. No integrated citator or subsequent treatment process is identified for judicial authorities, and the general requirement to verify output does not describe such a function. A country selector and an accessible source passage cannot determine whether a cited judgment was overturned or limited. That verification remains separate from the contract and matter analysis described for LEXA.
Xakia manages an in house team's matters, contracts and spend rather than case law, so a citator sits outside the product. Its AI reviews and summarizes the customer's own contracts and extracts key terms, and its query tools work across the customer's own files. Keeping contract standards in step with changes in law is the legal team's work, and the AI addendum makes users responsible for checking that output is accurate and up to date.
Refusal and Uncertainty Behavior
What does the product do when the answer is not in the corpus?
The security exhibit mentions optional confidence indicators without identifying a current display, calibration or threshold. LEXA's agents can follow precise instructions and return the clauses behind findings, but no explicit no answer path is described when source material cannot support a legal conclusion. The stated prohibition on signing or approving concerns action authority, not whether the model abstains from an unsupported answer. General warnings about mistakes and required human review therefore do not establish runtime refusal behavior.
Xakia's key terms extraction returns each value with a confidence indication, the clause it came from and the exact text it relied on. A user reviews and corrects each value before clicking Apply, so nothing reaches the contract repository unchecked. Contract review flags deviations from the playbook for a lawyer to decide, and the generated query remains visible for correction and reuse. Xakia describes no path for declining a request when a document holds no answer, such as a key term the contract does not contain.
The AI addendum tells users to validate or ignore output using their own professional judgment, because it may not be accurate or up to date.
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 contained no entry identifying Lexzur, App4Legal or LEXA as the responsible tool as of October 16, 2026. Incidental matches involving names such as Alexander and Alexa concerned other subjects and did not identify this product. The absence of an identified case does not measure LEXA's accuracy or prove that generated legal analysis is reliable. Its internal matter and agreement work may never appear in a reported court dispute, and users retain responsibility for verifying material used in proceedings.
The AI Hallucination Cases database maintained by Damien Charlotin, which records court decisions worldwide that address hallucinated AI content and the tool involved where known, has no entry naming Xakia. Xakia's AI produces contract redlines, summaries and extracted terms from the customer's own documents rather than court filings.
Bar Guidance Alignment
Has the vendor engaged in public with the ethics opinions its buyers are bound by?
Lexzur's educational material refers generally to American Bar Association guidance, competence, client interests, confidentiality and the risks of biased AI output. Its current agreement independently requires human verification and excludes professional advice. No named ethics opinion is connected to a specific LEXA control or mapped across jurisdictions. The older educational article's broad regulatory discussion does not establish the current legal position in each country.
Selecting a jurisdiction for a response also does not determine whether a firm's client consent, candor and supervision obligations have been satisfied.
Xakia's AI addendum refers to professional duties in general terms. It requires the AI features to be used only by or under the direct supervision of legal practitioners in accordance with applicable professional conduct rules. It also bars automated decisions with legal or similarly significant effects on individuals without proper human review in compliance with applicable professional ethics, guidelines and rules.
Neither the addendum nor Xakia's site names a bar opinion, such as ABA Formal Opinion 512, or guidance from the Law Council of Australia or a state law society, in Xakia's home market. Xakia's in house hub covers evaluating matter management software and legal operations practice.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
PRACTICE tracks time and expenses, supports hourly, fixed fee and capped arrangements, and generates invoices from approved entries. Lexzur promotes time savings from LEXA and improved billing efficiency, including a law firm customer account. These administrative features do not establish an invoice record identifying which billable work used AI, the actual time saved or the appropriate client charge. General AI activity logs record another part of the workflow.
No fee guidance connects compressed drafting or review time to the amount a lawyer may bill or the disclosure owed to the client.
Xakia's spend module is the payer's record of outside counsel fees. Outside firms accept work, update matter details and submit invoices through the Connect portal, which they use at no charge, and the legal team tracks spend against matter budgets in any currency. Xakia sets no limits on the number of firms, invoices or spend managed. It takes no percentage of invoices, which it calls a clip of the ticket, so its revenue does not grow with a customer's legal spend.
Budget tracking compares forecasts with spend to reduce overspend. AI invoice review now checks billing rules and discrepancies in totals, tax and vendor details; reviewers waive or act on findings. How firms bill AI assisted work is left to the customer's own billing terms.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
Lexzur makes its terms, data addendum, security exhibit and Azure OpenAI identification available for diligence. The full subprocessor list is supplied on request, and the trust center requires access for detailed resources. Those materials can support an initial explanation of the platform but do not provide a complete, immediately usable client disclosure package. Model versions, AI processing locations, retention exceptions and the difference between marketing and contractual improvement rights still affect an organization's assessment. No dedicated workflow applies a client's AI restrictions across LEXA features.
Xakia supplies a processor schedule identifying the country and purpose of providers including ChurnZero, HubSpot, Intercom, Microsoft, Stripe, Twilio and Xero. The DPA, AI addendum and security controls are available for diligence, while the ISO certificate, penetration test attestation and SOC 2 report require a request. No AI model provider is identified in the processor schedule despite third party model use under the addendum.
Matter restrictions and the limited Connect view control disclosure to external firms, but do not by themselves establish compliance with every client's AI requirements.
Court Disclosure Support
If a judge’s standing order requires an AI disclosure, can the product produce one?
LEXA retains chat history and states that AI actions are logged against the relevant record. The System Audit Log spans contracts, matters and tasks, with user, time, action and changed field information; filtered results can be exported to Excel. Contract activity includes approvals, signatures and negotiations. These features preserve useful portions of the work history, but no combined export ties every generated passage to the exact model version, retrieved sources and a lawyer's independent verification. A court declaration or certification therefore requires information beyond the described activity report.
Xakia keeps parts of the record a court or client might ask for. Extracted key terms carry the clause they came from and the exact text relied on, and reach the contract repository only after a user reviews them and clicks Apply. Xakia's AI page says each decision a user takes on an invoice review finding lands in the audit trail. The platform also records matter activity, intake routing and spend against each matter.
No export ties an AI output to the model used, and no model is named. Xakia's output is internal records, contracts and reporting rather than court filings, so a judicial order on AI disclosure is less likely to reach it.
The questions both sides leave open
Derived from the records above rather than written, so it cannot favor either vendor. Take these into both conversations and ask each side the same question.
- Good Law Verification
Which one fits
Choose Lexzur if
- Law firm billing belongs within the matter system. PRACTICE includes approved time and expenses, client invoices, fixed fees and capped arrangements.
- Contract administration requires its own lifecycle workflow alongside matters. CONTRA covers intake, templates, negotiation, approval, signature and ongoing obligations.
- The deployment requires private service or customer servers. Lexzur offers those choices, with separate configuration for the selected hosting arrangement.
Choose Xakia if
- The buyer is an internal legal department paying outside counsel. Xakia centers on departmental work allocation, budgets, matters and legal spend.
- AI analysis of incoming legal invoices is part of the project. Xakia includes that function within its legal operations offering, with allowances tied to the plan.
- External law firms need to participate without a portal invoice percentage charge. XakiaConnect provides that collaboration model alongside the department subscription.
In summary
Lexzur
Lexzur connects matter administration, practice finances and agreement workflows through PRACTICE and CONTRA. For teams building repeatable AI processes, the AI Legal Index highlights LEXA's specialized agents, chosen reference files and sequential or routing execution. Approval, signing and final legal risk decisions remain with people.
Xakia
Xakia organizes work for internal legal teams, including intake, matters, tasks, contracts, budgets and spend. Its AI functions include agreement review and analysis of outside counsel invoices. Subscription tiers specify user rates and AI allowances, while XakiaConnect supports external firm participation without a separate invoice percentage charge.
Questions buyers ask
How do Lexzur and Xakia charge for AI usage?
Lexzur supplies fifty introductory LEXA credits per instance and then sells transaction bundles. For the additional cost of AI, the AI Legal Index identifies Xakia's $50 per user monthly charge on Advance and $30 on Professional. Enterprise includes capped AI, while All In includes uncapped AI subject to fair use. The base user price and AI entitlement therefore need to be read together.
Do both products offer customer server installation?
Lexzur offers installation on customer servers as well as cloud and private service arrangements. Xakia operates as an Azure service with regional hosting choices. Lexzur's server option should not be read as proof that every LEXA model runs entirely inside the customer's infrastructure. From the AI Legal Index, based on each vendor's own published materials as of October 16, 2026. No vendor pays for placement.
What are their public per user entry rates?
Lexzur's Basic cloud subscription starts at $9 per user per month billed annually, for three to five users. Xakia's current entry rate is $100 per user per month, with higher plans at $125, $155 and $230 and an annual discount of up to ten percent. The products differ in included scope, AI allowances and external participation. From the AI Legal Index, based on each vendor's own published materials as of October 16, 2026. No vendor pays for placement.
Are Xakia's MCP connections part of the current comparison?
Xakia describes Claude and Copilot MCP connections as coming soon, so they are not treated as available functionality here. Its live AI invoice and agreement capabilities are separate. Lexzur's comparison rests on its current agent configuration and existing workflow connections. From the AI Legal Index, based on each vendor's own published materials as of October 16, 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.
Lexzur's entry subscription is for an individual product and a limited number of users. Xakia's forthcoming MCP connections are excluded from available capabilities. Neither vendor reviewed this page.