AI Agents & Assistants
Leena AI
Provenance unknown Report an error0–5 in half steps. 5 means the rubric's top anchor is met on the evidence.
by Leena AI Technologies Inc. · leena.ai
Report an error on this page Is this your product? →
Link to this page
Add ?theme=dark to the image address for dark backgrounds.
The badge shows the category and the month of the latest evaluation, and EU-Made where
that is the product’s provenance. It never shows a score.
Leena AI at a glance
- Vendor
- Leena AI Technologies Inc.
- Headquarters
- Not stated on the vendor’s pages as of 3 Nov 2026
- Data hosting
- Not stated on the vendor’s pages as of 3 Nov 2026
- Free plan
- Not stated on the vendor’s pages as of 3 Nov 2026
- Panel verdict
- The AI panel rates Leena AI 3/5, the mean of 6 judges. Leena AI is an agentic AI platform with pre-built AI Colleagues for back-office work. how scores are made
- As of
- Vendor pages captured 3 Nov 2026; evaluated by the panel 3 Nov 2026. all sources
Read this page as one judge. Each weighs the same scores by what they care about.
The panel's verdict
Leena AI is an agentic AI platform with pre-built AI Colleagues for back-office work. The bench scores it strongest on integrations and knowledge at a consistent 8 and on agent capabilities at 8-9, crediting 200+ pre-built integrations across 27 categories, MCP and A2A built in, high-risk actions on record updates, provisioning and money movement behind explicit approvals, and test cases saved per workflow version. Admin and access scored a consistent 7 for directory sync and permissions inherited from source systems. It is weakest on data use and training at 0-2: we found no public information on whether prompts, files or outputs train any model, on retention periods, or on a data processing agreement. Sovereignty scores span 2-4 over the same evidence — Netherlands governing law, unnamed 14+ deployment regions, and model providers named as OpenAI, Azure OpenAI, Anthropic and Amazon Bedrock with OpenAI GPT 5.5 as the default. Processing and models span 4-6, higher scores crediting the published model matrix, lower ones weighing the absence of a named EU processing option. We found no public pricing on the captured pages.
Speaks for it
- 200+ pre-built integrations across 27 categories, with documented end-to-end scopes in ServiceNow ticket lifecycle and Workday approvals.
- High-risk actions on record updates, provisioning and money movement sit behind explicit approvals, with guardrails blocking out-of-policy actions at runtime and violations logged as their own step in Run History.
- Every model and provider is published — OpenAI, Azure OpenAI, Anthropic and Amazon Bedrock, defaulting to OpenAI GPT 5.5 — with the resolved provider and model recorded for every run.
- Permissions are inherited from source systems with no second permissions table, under role-based controls reaching the individual skill.
- Test cases are saved per workflow version and can be run against the published version.
Held against it
- We found no public information on whether prompts, files or outputs train any model, on retention periods, or on a data processing agreement covering the product.
- We found no public information naming an EU processing option, per-provider inference regions, or a bring-your-own-key arrangement.
- We found no public information on the contracting entity's jurisdiction, the hosting location, or a subprocessor list.
- We found no public information naming a SAML/OIDC single sign-on flow, SCIM provisioning or SIEM export of audit logs, with single sign-on appearing only as an integration category.
- We found no public statement on human handover carrying full conversation context, on per-agent quality and cost monitoring, or on the vendor's own account of what these agents must not be trusted with.
Best for
- You need agents that complete actions end to end in systems such as ServiceNow and Workday, with money movement and provisioning gated behind explicit approvals.
- Your team wants model choice and traceability, with a published model matrix and the resolved provider and model recorded for every run.
- You govern access through Active Directory, your HRIS and source-system permissions, and need role-based control down to the individual skill with a full audit trail.
- Your stack includes systems without an API, and you value browser automation and a governed no-code REST connector alongside the pre-built integrations.
Avoid if
- Your procurement requires a data processing agreement and a published position on training use and retention before rollout; the captured terms address cookies and free-website liability.
- You require European model providers or a stated EU processing region; the named providers are OpenAI, Azure OpenAI, Anthropic and Amazon Bedrock, with OpenAI GPT 5.5 as the default configuration.
- Your security team needs SAML/OIDC single sign-on, SCIM provisioning and SIEM log export documented in writing; single sign-on appears in the pages only as an integration category.
- A works council must approve a platform whose full trace log records every employee conversation; we found no public information on how long those logs are kept.
The scores
| Criterion | Panel median |
|---|---|
| Data use, training & retention | 0.5 of 5 |
| Models & processing location | 2 of 5 |
| Admin, identity & access controls | 3.5 of 5 |
| Integrations, connectors & knowledge | 4 of 5 |
| Agent capabilities & guardrails | 4 of 5 |
| European sovereignty | 1.5 of 5 |
Data use, training & retention
n/r
n/r
Show reasoningHide reasoning
How this is scored
What happens to prompts, uploaded files, connected data and outputs: whether any of it trains a model (the vendor's or a model provider's), how long it is kept and where logs live, whether zero-retention is available, and whether a data processing agreement covers it.
0 — No statement on training or retention; the privacy policy is a website policy that does not cover the product.
3 — A general statement that business data is "not used for training", with no word on retention, on the model providers behind the product or on logs.
5 — Business data excluded from training by contract, retention periods stated for chats and files, and a DPA offered to business customers.
8 — No-training commitments stated for the vendor and every model provider it routes to, retention configurable by the admin (including deletion of chat history), zero-data-retention terms with the model providers named, and abuse-monitoring retention described.
10 — Data use fully accounted for: training, retention, logging and abuse monitoring stated per model provider, customer-controlled retention and deletion with an audit trail, outputs and connected data treated as customer data in the DPA, and an independent audit (SOC 2, ISO 27001 or ISO 42001) covering the product.
The Head of IT
I found no public statement on whether prompts, uploaded files or outputs train a model, how long chats and run logs are kept, or any zero-retention terms with the model providers behind the platform. The published certifications (SOC 2 Type II, ISO 27701, GDPR) and documented encryption at rest and in transit speak to security posture, not to data-use commitments, and I would not sign eight hundred staff onto this on the evidence captured. 2 4
The Data Protection Officer
I found no statement on whether prompts, files, connected data or outputs train any model, on retention periods, on zero-retention terms, or on a data processing agreement covering the product; the captured terms cover website cookies and a liability disclaimer for free website use. Encryption at rest and in transit, a full trace log and SOC 2 / ISO certifications are documented, and PII in conversations can be masked or blocked — but none of that answers where a prompt is kept or who learns from it. 2 4 7
The Automation Lead not rated — no public information
We found no public information on whether prompts, uploaded files or outputs train any model, how long chats, files and logs are kept, or whether a data processing agreement and zero-retention terms cover the product. The captured pages document encryption, certifications and guardrails that mask PII in conversations, but say nothing about training, retention periods or the model providers' data terms. 2 4
The Customer-Service Director
I found no public information on whether our conversations, uploaded files or connected ticket and HR data are used to train any model, nor on retention periods, zero-retention terms or a data processing agreement for the product. What is published is a security posture — SOC 2 Type II and ISO 27001/27701 certifications, AES-256-GCM at rest and TLS 1.2+ in transit — which my DPO will welcome but which does not answer the training and retention question. 2 4
The Works Council Member not rated — no public information
We found no public information on whether staff prompts, uploaded files or outputs train any model, how long conversations are kept, or whether a data processing agreement covers the product. For a tool that logs every employee conversation, that silence is exactly what a works council cannot sign off on; the captured terms and platform pages say nothing on training, retention or log storage. 2 4 3
The Skeptic
The terms cover cookies and free-website liability, and the platform pages state encryption and a certification list, but on whether prompts, files or outputs train any model, on retention periods, on logs, and on a data processing agreement we found no public information. For a platform whose agents touch payroll and money movement, that silence is the first problem. 2 4
Models & processing location
Show reasoningHide reasoning
How this is scored
Which models sit behind the product and who runs them, where inference and storage happen, whether the customer can pin processing to the EU, choose or exclude model providers, bring its own model or key, or run the product in its own cloud or on premises.
0 — The models behind the product and the processing location are not stated.
3 — The model family is named, but where inference runs and through which provider (the model maker, a hyperscaler, the vendor) is unstated.
5 — Models and their providers named, with an EU processing option stated for storage and at least the default model.
8 — Each available model listed with provider and inference region, EU-only processing selectable by the admin and enforced across models, model providers excludable, and bring-your-own-key or own-cloud deployment documented.
10 — Processing fully under the customer's control and documented: EU-only or on-premises inference for every feature including agents and connectors, open-weight or self-hosted models supported, data flows per feature published, and a residency commitment in the contract rather than on a marketing page.
The Head of IT
The model layer is documented better than most: OpenAI, Azure OpenAI, Anthropic and Amazon Bedrock are each listed with their models, the default is OpenAI GPT 5.5, effort and priority settings, retirement dates and automatic migrations are stated, and every run records which provider and model handled it. Deployment offers public cloud, single-tenant or private VPC across 14+ regions framed around data residency, but I found no public information naming per-model inference regions, an enforced EU-only option, provider exclusion or bring-your-own-key. 3 4
The Data Protection Officer
The model matrix is unusually public: OpenAI, Azure OpenAI, Anthropic and Amazon Bedrock are each listed with their models, the default is OpenAI GPT 5.5, and retiring models have published replacement paths. Deployment offers public cloud, single-tenant or private VPC across fourteen-plus regions; what I found no statement on is an EU processing option for storage or inference, the region each provider runs in, or any way to pin or exclude a provider. 3 4
The Automation Lead
Every model is listed by provider — OpenAI, Azure OpenAI, Anthropic and Amazon Bedrock — configurable per orchestrator with inheritance and override, with run history recording which provider and model handled each request. Deployment comes as public cloud, single-tenant or private VPC across 14+ regions framed around data residency, but we found no public information pinning inference or storage to a named EU region, on bring-your-own-key, or on where each provider runs inference. 3 4
The Customer-Service Director
Every model and provider is documented — OpenAI, Azure OpenAI, Anthropic and Amazon Bedrock, defaulting to OpenAI GPT 5.5 — with per-agent override or live inheritance from the master orchestrator and a run-history trace of which model handled each request. Deployment offers public cloud, single-tenant or a private VPC across 14+ regions, but an explicit EU processing choice and the inference region per provider are not stated, which is what I would need pinned before go-live. 3 4
The Works Council Member
The documentation names every model and its provider — OpenAI, Azure OpenAI, Anthropic and Amazon Bedrock — and lets the administrator set the model for the orchestrator and each agent, which we welcome. But the captured pages state neither where inference runs nor an EU processing option; deployment is advertised only as public cloud, single-tenant or private VDC across '14+ deployment regions' with not one region named, so staff prompts may leave Europe with nothing published to check. 3 4
The Skeptic
Credit where due: the docs name every model and provider — OpenAI GPT 5.5 as the default, plus Azure OpenAI, Anthropic and Amazon Bedrock — and record the resolved provider and model for every run. Where inference physically runs is unstated; deployment options mention single-tenant and private VPC across fourteen-plus regions to fit data residency requirements, but no EU processing option is named, and the pages say nothing about bring-your-own-key or on-premises running. 3 4
Admin, identity & access controls
Show reasoningHide reasoning
How this is scored
What the IT department controls: SSO and SCIM provisioning, roles and groups, which models, agents and connectors each group may use, retrieval that respects the permissions of the source system, audit logs, usage reporting, and policies such as data-loss prevention — and whether reporting on staff use is scoped and documented.
0 — No admin features documented; accounts are individual sign-ups.
3 — A team workspace with an owner and member invitations, but no SSO, no roles and no audit log.
5 — SSO, admin and member roles, user management, and an admin view of usage.
8 — SAML/OIDC SSO with SCIM provisioning, group-based permissions for models, agents and connectors, retrieval that inherits source-system permissions, exportable audit logs, and usage analytics with documented scope.
10 — Governance a regulated company can operate: fine-grained policies (DLP, content filters, prompt and output logging rules), audit logs streamed to a SIEM, per-agent ownership and approval, usage reporting that can be aggregated or pseudonymised for works-council agreements, and every admin action logged.
The Head of IT
Directory sync from Active Directory and the HRIS, role-based control at the colleague, sub-agent and individual-skill level, and permissions inherited from the source system with no second table to drift are exactly what I need to govern who gets which model and agent. A full audit trail, per-user attribution of every action and an Identity & SSO integration category are documented; I found no public word on SCIM, SAML/OIDC specifics, SIEM streaming of logs, or the scope of usage reporting. 1 4 7
The Data Protection Officer
Role-based control reaches the individual skill, directory sync from Active Directory and the HRIS is documented, retrieval inherits the source system's permissions with no second permissions table, every change is written to the audit trail, and MCP calls are tracked in analytics. I found no naming of SAML/OIDC or SCIM, no export of audit logs to a SIEM, and the scoping of usage reporting for works-council purposes is not described — an identity and SSO connector category exists, but as a category name, not a documented control. 4 5 6 7
The Automation Lead
Directory sync from Active Directory and the HRIS feeds role-based controls at the AI Colleague, orchestrator and individual skill level, permissions stay in the source system so retrieval inherits them, and every change is written to an audit trail with a full trace log and per-user action attribution. We found no public information naming SAML/OIDC single sign-on with SCIM provisioning, exportable logs or SIEM streaming, which is what keeps this below the top tier. 4 7 6
The Customer-Service Director
Directory sync from Active Directory and the HRIS, role-based controls at the agent, sub-agent and individual-skill level, permissions that stay in the source system with no second table to drift, tenant isolation, a full trace log and every action attributable to the user it was taken for — that is real governance for a forty-seat operation. I found no public information on SCIM provisioning, on exporting audit logs to a SIEM, or on how usage reporting on staff is scoped. 4 5 1
The Works Council Member
Directory sync from Active Directory and the HRIS, role-based controls at the agent and individual-skill level, permissions inherited from the source systems and a full trace log with every action attributable to a user — retrieval respecting source permissions is documented and matters to us. What we could not find is the works-council piece: beyond an 'Identity & SSO' integration category, we found no public information on SSO mechanism or SCIM provisioning, on exporting audit logs to a SIEM, or on whether usage reporting on staff can be aggregated or pseudonymised. 4 1 6 5
The Skeptic
Real IT-grade material: directory sync from Active Directory and HRIS, role-based controls at the AI Colleague, sub-agent and individual skill level, permissions that stay in the source system rather than in a second table, tenant isolation, and an audit trail written on every change. The pages name no SAML/OIDC or SCIM flow, and we found no public information on exportable audit logs or on usage analytics with documented scope. 1 4 7
Integrations, connectors & knowledge
Show reasoningHide reasoning
How this is scored
How the product reaches company knowledge and systems: connectors to file stores, wikis, mail, CRM, ticketing and ERP, how fast and how completely they sync, the API, support for open protocols such as MCP, and how the assistant cites the sources it answers from.
0 — No integrations; users paste text or upload files one at a time.
3 — File upload and one or two connectors (for example a drive), with no API and no statement on sync or citations.
5 — A connector library covering the common file stores, wikis and collaboration tools, answers citing sources, and a documented API.
8 — Connectors to file stores, wikis, mail, CRM and ticketing with incremental sync, custom connectors or MCP support, an API and webhooks for embedding assistants and agents in other tools, and source citations on every grounded answer.
10 — The product sits in the company's systems: read and write connectors with documented scopes, MCP servers and clients, contact-centre or telephony integration where relevant, a knowledge layer that can be audited and corrected, and the data, agents and configuration exportable when the customer leaves.
The Head of IT
Two hundred pre-built enterprise integrations across twenty-seven categories — Workday, SAP, ServiceNow, Salesforce, Oracle — plus a thousand pre-built tools, a no-code REST connector, browser automation for systems without APIs, SFTP and LDAP give me the reach my stack needs. MCP and A2A are built in, MCP calls carry real user identity, stay permission-aware and return grounded answers with source links; I found no public statement on sync frequency or on exporting data, agents and configuration when leaving. 1 5 6
The Data Protection Officer
Two hundred plus pre-built connectors across twenty-seven categories, deep documented scopes for Workday and ServiceNow, MCP on both client and server side with permission-aware calls and source links on grounded answers, A2A handoffs, a governed no-code REST connector, SFTP, LDAP and browser automation. I found no statements on connector sync frequency, on webhooks, or on exporting data, agents and configuration when a customer leaves. 1 5 6
The Automation Lead
This is the deep part for me: 200+ pre-built integrations across 27 categories including ERP, CRM, ITSM and knowledge systems, documented write actions in ServiceNow and Workday, MCP built in both as a client tool and a callable server, agent-to-agent handoffs, a REST API fallback, SFTP, LDAP and governed browser automation for systems with no API, and grounded answers carrying source links. We found no public information on sync frequency, webhooks, or export of data and configuration when a customer leaves. 5 6 1
The Customer-Service Director
Two hundred-plus pre-built integrations across twenty-seven categories including ITSM, CRM, messaging and knowledge systems, a thousand pre-built tools, ServiceNow ticket lifecycle end to end, native MCP and A2A with a REST API fallback, a no-code connector for any REST API, and answers grounded with source links. Sync frequency and webhooks are not documented, which is the only reason I hold it here — the ticketing depth is exactly what my desk needs. 1 5 6
The Works Council Member
Two hundred pre-built enterprise integrations across HR, ITSM, ERP, CRM and knowledge systems, MCP built in with permission-aware calls grounded with source links, a governed no-code REST connector, LDAP, SFTP and even browser automation for systems without an API. We found no public information on sync cadence or on exporting data, agents and connectors when the customer leaves. 5 6 1
The Skeptic
Two hundred-plus pre-built integrations across twenty-seven categories with concrete scopes — ServiceNow tickets raised, updated and resolved end to end, Workday leave and lifecycle approvals — plus MCP built in as both server and client with a REST API fallback, a governed no-code REST connector, LDAP, SFTP and browser automation for systems without an API, with MCP answers permission-aware and grounded with source links. We found no public information on sync speed or completeness, or on exporting data, agents and configuration when a customer leaves. 1 5 6
Agent capabilities & guardrails
Show reasoningHide reasoning
How this is scored
What an agent can actually do and where it stops: multi-step tasks, tool use and actions in other systems, scheduling and triggers, human approval before consequential actions, run logs, evaluation and testing before release, and the limits the vendor states rather than hides.
0 — No agents; the product is a chat window, whatever the marketing says.
3 — Saved prompts or custom assistants with instructions and files, but no actions in other systems and no run history.
5 — Agents that call tools or APIs and run multi-step tasks, built without code, with a log of each run.
8 — Agents that act in connected systems with scoped permissions, human approval configurable per action, triggers and schedules, test and evaluation tooling before release, and documented limits and failure handling.
10 — Agents a business can put in production and account for: versioning and staged rollout, evaluation against test sets, monitoring of quality and cost per agent, handover to a human with full context, every action attributable to the agent and the person who deployed it, and the vendor's own statement of what agents must not be trusted with.
The Head of IT
Agents work end to end in connected systems (raising, updating and resolving tickets; leave and lifecycle changes in Workday), high-risk actions — record updates, provisioning, money movement — sit behind explicit approvals, and every action is attributable to the user it was taken for. Guardrails are layered and enforced at runtime with violations visible as their own step in run history, and test cases are saved per workflow version with a view-only mode for the published version; I found no public statement on staged rollout, per-agent cost monitoring, or the vendor's own account of what agents must not be trusted with. 4 5 7 8
The Data Protection Officer
High-risk actions — record updates, provisioning, money movement — sit behind explicit approvals, four layers of guardrails block out-of-policy actions before execution with violations logged per turn in run history, every action is attributable to the user it was taken for, and workflow versions carry versioned test cases with published-version test runs and documented failure handling. I found no public statement on staged rollout, per-agent cost monitoring, human handover with full context, or the vendor's own account of what these agents must not be trusted with. 3 4 7 8
The Automation Lead
Agents act in connected systems with scoped permissions, and high-risk actions — record updates, provisioning, money movement — sit behind explicit approvals, which is exactly what I need before an agent touches the ERP. Guardrails are enforced at four layers with runtime blocking rather than after-the-fact flagging, violations appear as their own step in run history, every action is attributable to a user, and test cases are saved per workflow version and can be run against the published version before rollout. We found no public information on per-agent quality and cost monitoring, human handover with full context, or a vendor statement of what agents must not be trusted with. 4 7 8
The Customer-Service Director
Agents raise, update and resolve ServiceNow tickets and run Workday approvals end to end, high-risk actions on record updates, provisioning and money movement sit behind explicit approvals, out-of-policy actions are blocked rather than flagged after the fact, run history shows which model and which guardrail fired, and test cases are saved per workflow version and runnable against the published version — genuine pre-go-live evaluation. My one deal-breaker gap: I found no public information on handover to a human agent carrying the full conversation context, and no monitoring of quality or cost per agent. 4 5 7 8
The Works Council Member
High-risk actions — record updates, provisioning, money movement — sit behind explicit approvals, guardrails validate both what employees send and what the AI answers and block before execution, every action is attributable to the user it was taken for, and each run is logged with the resolved model. Test cases are saved per workflow version and can be run against the published version, which is real pre-release discipline; what we miss is monitoring of quality and cost per agent and the vendor's own statement of what these colleagues must not be trusted with. 4 7 8 3 5
The Skeptic
Agents act in connected systems on validated fields, high-risk actions — record updates, provisioning, money movement — sit behind explicit approvals, guardrails block out-of-policy actions at runtime with each violation logged as its own step in Run History, test cases are saved per workflow version and can run against the published version, and every action is attributable to the user it was taken for. The pages are silent on schedules, on monitoring quality and cost per agent, and on the vendor's own statement of what these agents must not be trusted with. 3 4 7 8
European sovereignty
panel opinion
Show reasoningHide reasoning
How this is scored
Who the contracting entity and the parent company are, where customer data is hosted, and who the subprocessors are — model providers above all, then hosting, search, telephony and email. Independently sourced by the sovereignty pipeline; for an AI product the model provider is a subprocessor, so a European interface over a non-European model counts as such.
0 — Non-EU vendor and contracting entity, hosting unstated, subprocessors (including model providers) unnamed.
3 — A privacy policy exists, but hosting location is unstated or non-EU and the model providers behind the product are not named as subprocessors.
5 — EU hosting stated and an EU contracting entity, but prompts reach non-EU model providers or other subprocessors without a stated safeguard or EU region.
8 — EU hosting on named infrastructure, EU contracting entity, every subprocessor listed with its location including model providers, EU inference available, and a DPA offered.
10 — Sovereign end to end and evidenced: European ownership, entity, hosting and model processing, every subprocessor European or optional, data export and deletion documented, and a security or certification statement published rather than implied.
The Head of IT
The contracting entity is Leena AI Technologies Inc. and every sovereignty dimension I need — jurisdiction, hosting, data-processing agreement, transfer safeguards — is reported as unknown in what was captured; the website terms name Netherlands governing law, but that governs a free-to-use website rather than the product's data processing. The model providers named behind the product are all non-European (OpenAI, Anthropic, Amazon Bedrock), and I found no public subprocessor list, no stated EU hosting for customer data and no data-processing agreement, only a GDPR compliance claim, certifications and multi-region deployment language. 2 3 4
The Data Protection Officer
A privacy policy exists and the terms are governed by the law of the Netherlands, but the contracting entity's jurisdiction is not published, hosting is stated only as fourteen-plus unnamed regions, and the model providers — OpenAI, Azure OpenAI, Anthropic, Amazon Bedrock — are named in the configuration documentation while I found no subprocessor list giving their locations or regions. No EU hosting, EU inference or residency commitment is stated in anything I read; as it stands, I cannot tell where a prompt physically travels or under which processor's terms. 2 3 4
The Automation Lead
The model layer is named — OpenAI, Azure OpenAI, Anthropic and Amazon Bedrock, with OpenAI GPT 5.5 as the default configuration — and deployment spans 14+ regions in public cloud, single-tenant or private VPC, but we found no public statement of EU hosting, a named EU region, an EU contracting entity or a subprocessor list. The terms name Netherlands governing law while the vendor is an Inc. and the contracting entity's jurisdiction is unstated. 3 4 2
The Customer-Service Director
Dutch governing law on the terms and a choice of public cloud, single-tenant or a private VPC across 14+ deployment regions are the positives I can verify. The hosting location for our data is not stated, the models behind the product are OpenAI, Anthropic and Amazon Bedrock — all non-European — and I found no public information on a subprocessor list, a stated EU-region safeguard, or a data processing agreement. 2 3 4
The Works Council Member
The models behind the product are OpenAI, Anthropic and Amazon Bedrock — all non-European — while the terms run under Netherlands law for a vendor trading as an Inc., and hosting is offered as public cloud, single-tenant or private VPC across unnamed '14+ deployment regions'. No subprocessor list or data processing agreement appears in the captured pages, so the safeguards carrying staff prompts to those providers are unpublished even though SOC 2, ISO 27001 and GDPR are claimed as certifications. 3 2 4
The Skeptic
A Dutch governing-law clause in the terms, and then OpenAI GPT 5.5 as the default configuration with Anthropic and Amazon Bedrock behind it — non-EU model providers in the routing table. The contracting entity's jurisdiction, the hosting location and a data processing agreement are unknown on every dimension I can check, and we found no EU hosting or EU inference commitment on any page. 2 3 4
Pricing transparency
not rated — the vendor publishes no price
Show reasoningHide reasoning
How this is scored
Whether a company can compute the real annual cost from public pages alone: price per seat, per agent, per conversation or per usage unit, what is included in each tier, usage limits and overage, which models, connectors or agents cost extra, minimum seats and term.
0 — No public prices at all; every customer is a sales call.
3 — A starting price or "from" figure exists, but usage limits, included models and what scales the cost are unclear.
5 — Prices per tier are public with the main inclusions, but usage limits, overage, agent runs or enterprise features (SSO, audit logs, EU processing) are unpriced or "on request".
8 — Every tier priced publicly with seat or usage units, usage limits and overage rates, add-ons listed, minimum seats and term stated.
10 — Complete price computability: annual cost derivable for a given number of seats, agents or conversations, with every usage rate, model surcharge, add-on and support tier and the cancellation and data-return terms published.
The Head of IT not rated — no public information
We found no public pricing information on the pages captured — no per-seat price, no tiers, no usage limits, no minimum seats or term. The homepage and integrations pages carry outcome claims instead (measured ROI of 4–10×, hard-dollar savings quoted as $1M–$10M+ per year, payback in about three months), none of which tells me anything I could put in a budget. 1 2 5
The Data Protection Officer not rated — no public information
We found no public information on pricing: no per-seat, per-agent or usage prices, no tiers, usage limits, overage, minimum seats or term appear on the captured pages. The homepage publishes ROI and savings ranges, which are outcomes rather than prices. 1 2
The Automation Lead not rated — no public information
The captured pages carry ROI claims — payback in one quarter at roughly three months, measured ROI of four to ten times — but we found no public information on pricing itself: no seat or usage price, no tier inclusions, no usage limits, no term or minimum. A buyer cannot compute an annual cost from these pages. 1 2
The Customer-Service Director not rated — no public information
I found no public information on pricing anywhere I read — no per-seat, per-agent or per-conversation price, no tiers, usage limits or term on the homepage, the terms, the platform, the integrations or the docs pages. The only cost-related figures published are savings and ROI claims, and my CFO cannot turn 4–10× measured ROI into a forecast of cost per conversation. 1 2 4 5 6
The Works Council Member not rated — no public information
We found no public information on pricing: no seat or usage price, tier, usage limit, minimum term or cancellation term appears anywhere in the captured pages — only ROI claims such as 'payback in one quarter' and savings ranges. A council cannot budget a deflection figure. 1 2
European sovereignty — proven facts
0 of 6 dimensions provenBuilt only from facts shown on the vendor's own pages. A dimension we could not prove is left open, not scored as zero.
| Legal entity | Not determined | — | uncited Report an error |
|---|---|---|---|
| Hosting location | Not determined | — | uncited Report an error |
| Hosting options | Not determined | — | uncited Report an error |
| Open source | Not determined | — | uncited Report an error |
| Data processing agreement | Not determined | — | uncited Report an error |
| Transfer impact assessment | Not determined | — | uncited Report an error |
Questions about Leena AI
Is Leena AI European?
Not determined: the vendor’s pages as captured do not say where Leena AI is based.
Where does Leena AI store data?
Not stated on the vendor’s pages as of 3 Nov 2026.
Does Leena AI have a free plan?
Not stated on the vendor’s pages as of 3 Nov 2026.
What are the best alternatives to Leena AI?
As of 4 Nov 2026, the panel rates Langdock (4/5), Dust (3.5/5) and Unique AI (3.5/5) highest among the other 18 AI Agents & Assistants products evaluated. Ratings that look equal are rounded to half stars; the order uses the unrounded rating, then the name. The highest-rated European alternative is Langdock (EU-Made). Ratings are the AI panel’s opinion, formed from public evidence: how scores are made. The full list, with European products first: Leena AI alternatives.
Where this could be wrong
- Evidence ages. The oldest capture behind this page is from 3 Nov 2026. Vendors change pricing and policies without notice; every fact reflects its source as of the capture date shown in the registry.
- AI can misread a source. Extraction and judgement are automated; a citation guarantees traceability, not infallibility. If something here is wrong, say so — no account needed, every report is decided within 5 business days, and accepted corrections are published.
What we left out
A claim that does not survive our checks costs us the claim, not the page. This is what was taken off this one.
- We could not confirm any pricing information on the vendor’s own pages as captured, so this page shows none rather than a statement we cannot stand behind. Know more? Tell us
- 19 product facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material. Know more? Tell us
- 15 integrations facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material. Know more? Tell us
- 6 compliance facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material. Know more? Tell us
- 6 support facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material. Know more? Tell us
- 5 legal facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material. Know more? Tell us
- 1 hosting fact could not be confirmed on the vendor’s page as captured and was left out of this page and of the panel’s material. Know more? Tell us
Sources (8)
The pages every claim on this page was read from — each one checked, dated, and kept verifiable.
- 1 Vendor page leena.ai Checked 3 Nov 2026 Details →
- 2 Terms of service — found from the homepage leena.ai Checked 3 Nov 2026 Details →
- 3 Models & processing location — found from sitemap docs.leena.ai Checked 3 Nov 2026 Details →
- 4 Admin, identity & access controls — found from sitemap leena.ai Checked 3 Nov 2026 Details →
- 5 Integrations, connectors & knowledge — found from sitemap leena.ai Checked 3 Nov 2026 Details →
- 6 Integrations, connectors & knowledge — found from sitemap leena.ai Checked 3 Nov 2026 Details →
- 7 Agent capabilities & guardrails — found from sitemap docs.leena.ai Checked 3 Nov 2026 Details →
- 8 Agent capabilities & guardrails — found from sitemap docs.leena.ai Checked 3 Nov 2026 Details →