Select a name to see it on the map. Open a profile to see every fact and source.
What the source marks mean Each mark shows what kind of source backs a fact. Open the fact to read the source, the date it was checked and any conditions.
The provider's own documentation supports this fact, for the product, plan and date stated.
The provider says this, for example on a marketing page. Read the linked source and its conditions before you rely on it. ETHORITY's assessment, based on the linked sources.
An independent source supports this one fact, within the scope stated. Check the other facts separately. Figures marked 'Scenario' are calculated from stated assumptions. Results measured in practice are shown separately.
Documents, Workflow & Agents Dify is a source-available platform for visually building LLM chat apps, agents and workflows with knowledge bases, Marketplace model and tool plugins, human-input pause nodes, run cloud-hosted or self-hosted.
Per workspace /mo ○ Conditions apply – open for details Your region (Ent.) ○ Conditions apply – open for details LD Dify
LangGenius (Dify) · Platform
Facts, sources and conditions Dify is a source-available platform for visually building LLM chat apps, agents and workflows with knowledge bases, Marketplace model and tool plugins, human-input pause nodes, run cloud-hosted or self-hosted.
Where it sits on the map · 1 · 1 Build and run LLM agents and workflows (cloud or self-hosted)
Task classification by ETHORITY; match it to your intended use.
Per workspace per month (monthly or annual), with message-credit, member, app and knowledge quotas
Enterprise: contact sales, no price
Sandbox free (200 message credits); Professional $59/workspace/month or $590/year (3 members, 5,000 credits/month); Team $159/workspace/month or $1,590/year (50 members, 10,000 credits/month); taxes extra
Unit: USD (shown as $) currency code not named on page
Dify Cloud region not named ("in our managed cloud region"); Enterprise: Dify-managed single-tenant cluster "in your region" or customer-managed in own AWS/GCP/Azure account; privacy policy lists processing in US, Australia, Canada, China, EEA and UK
Cloud wording from https://dify.ai/pricing/dify-cloud; processing locations from https://dify.ai/legal/privacy-policy (website servers 'located in the United States'). No named EU region for Dify Cloud
Not used for training: "We do not train AI models ourselves, and we will not use your AI interaction data for model training."
With own API key, training depends on customer's agreement with that model provider; sub-processors 'prohibited from using your data for model training'
Model providers, tools, data sources and MCP integrations installable from the Dify Marketplace or self-built
SOC 2 Type II (attestation, auditor Sensiba); ISO 27001:2022 (assessed by Johanson); scope boundary not stated
Where the provider's own sources disagree The provider’s own sources say different things here. The trust centre names SOC 2 Type 1; the March 2026 blog and enterprise page name Type II. Confirm the applicable report.
Human Input web-app delivery not available in Trigger-started workflows; request closes after first response; default 3-day timeout; upload caps (e.g. 15 MB documents); Sandbox log history 30 days
plan limits from https://dify.ai/pricing/dify-cloud
DR Dropbox Dash Dropbox · Product
According to Dropbox: “Dash searches across apps like OneDrive, Notion, and Gmail—finding files, images, or videos fast with AI-powered universal search.”
Per user per month ● Conditions apply – open for details US only ● Conditions apply – open for details DR Dropbox Dash
Dropbox · Product
Facts, sources and conditions According to Dropbox: “Dash searches across apps like OneDrive, Notion, and Gmail—finding files, images, or videos fast with AI-powered universal search.”
Where it sits on the map · 1 · 1 AI universal search and chat across connected work apps
Derived from plans page and connector help article.
Per user per month (yearly or monthly billing)
Business plan yearly only. Possible future per-member item limit with 30 days' notice.
Dash for Teams USD 15/user/month yearly or USD 19 monthly; Dash for Business USD 35/user/month yearly only; 50% off for Dropbox Standard/Advanced/Business/Business Plus customers
Unit: USD Only USD shown; no EUR prices on page.
No: Dash data stored only in the US; no option for storage outside the US; EU transfers under SCCs and EU-U.S. Data Privacy Framework
Dash Security Whitepaper v2026.3 specifies US-only data storage; confirm the region for your account.
Only with opt-in
Quote: 'We will not build generative AI models using your content without consent.' Whitepaper: usage/interaction data may improve Dash (opt-out available); de-identified Q&A snippets may be manually reviewed; LLM partners 'don't train on or retain your data beyond their contractual retention window'. Plans page phrased without the consent qualifier.
25 apps: 16 personal (Airtable, Asana, GitHub, Gmail, Google Calendar, HubSpot, Lucid, Microsoft 365, OneDrive personal, Outlook, Miro, Monday, Notion, Smartsheet, Zendesk, Zoom) + 9 admin team apps (Confluence, Dropbox, Gong, Google Drive, Jira, OneDrive, Microsoft Teams, Slack, Workday)
Connector count follows the names listed in the source.
SOC 2 Type II (Security criteria) attestation for Dash; ISO/IEC 27001 certificate for Dash (scope not stated); auditor/issuer not named
Security page (https://dash.dropbox.com/security) mentions SOC 2 Type II but not ISO 27001. GDPR/CCPA/DPF on security page are laws/frameworks, not certifications. Not verified on issuer site.
US-only storage; content exclusions only for Google Drive and Microsoft 365; multiple profiles per app not supported; sync takes minutes to several days; no access to local files or browser history; some apps use team-wide permissions; automated policy fixes can take hours
Combined with Dash Security Whitepaper V2026.3.
Documents, Workflow & Agents Dust lets company teams build shared AI agents over connected knowledge sources and tools, choosing among 20+ models per task, with stake-level approvals before agents take actions like sending email.
Per seat/mo + credits ○ Conditions apply – open for details "Host in the EU or US t… ○ Conditions apply – open for details Facts, sources and conditions Dust lets company teams build shared AI agents over connected knowledge sources and tools, choosing among 20+ models per task, with stake-level approvals before agents take actions like sending email.
Gmail action approvals follow the tool guide; the model count follows the pricing page. Confirm settings for each enabled tool.
Where it sits on the map · 1 · 1 Build team AI agents over company knowledge and tools
Task classification by ETHORITY; match it to your intended use.
Per seat per month with monthly credit allowance (Pro 8,000, Max 40,000 credits per seat/month); programmatic usage $0.01 per credit; Enterprise workspace-pooled credits
Credits reset each billing period, no rollover; overage capped by admin
Free €0 (500 lifetime credits); Pro €24 per seat/month billed yearly; Max €120 per seat/month billed yearly; excl. VAT; Enterprise on quote
Unit: EUR per seat per month Where the provider's own sources disagree The provider’s own sources say different things here. Business plans for teams up to 100 people; yearly toggle 'Save 20%'. Page meta description gives $24/$120 (currency inconsistency on same page)
"Host in the EU or US to meet your regulatory needs"; US and EU data residency on Business and Enterprise; single-tenant deployment Enterprise only
Ask which cloud region and hosting provider apply to your account.
Not used for training: "Your data is never used to train models."; "No data is stored by third-party model providers."
Enterprise page: "Zero model training on your data"
Connections, MCP servers, Open API; named: Notion, Slack, Salesforce, GitHub, Gmail, Outlook, Zendesk; homepage "70+ connectors"; Business plan up to 3 connectors and 5 remote MCP servers, Enterprise unlimited
Where the provider's own sources disagree The provider’s own sources say different things here. Inconsistent counts across own pages: pricing 'Connectors to 20+ sources' vs homepage '70+ connectors' (https://dust.tt/)
Provider claim: "GDPR Compliant & SOC2 Type II Certified. Enables HIPAA compliance." No auditor, scope or report period on pages read
SOC 2 is an attestation; GDPR and HIPAA are laws. Confirm the report's scope and period.
"Dust is not designed for bulk email campaigns"; "send_mail is disabled by default on existing workspaces"; Business plan capped at 3 connectors, 5 Spaces, teams up to 100
Plan caps from https://dust.tt/home/pricing
EL Elastic Workplace Search Elastic · Product
Elastic workplace search across all your tools, which teams combine with generative AI to build their own Slackbots, website portals and mobile employee experiences.
Usage-based · monthly ○ Conditions apply – open for details EU regions selectable ● Conditions apply – open for details EL Elastic Workplace Search
Elastic · Product
Facts, sources and conditions Elastic workplace search across all your tools, which teams combine with generative AI to build their own Slackbots, website portals and mobile employee experiences.
Where the provider's own sources disagree The provider’s own sources say different things here. The solution page still promotes Workplace Search with a trial, while the 15 April 2025 provider announcement says standalone App Search and Workplace Search are deprecated in 9.0, available in 8.x and supported for 27 months after the end-of-life announcement. Confirm the supported version and migration plan. https://www.elastic.co/blog/search-for-enterprise-serverless
Where it sits on the map · 1 · 1 Build-your-own enterprise search across workplace sources on Elasticsearch
Derived from solution page; standalone Workplace Search product deprecated, capability continues via Elasticsearch connectors.
Elastic Cloud Hosted: monthly, usage-based by instance type and subscription tier (Standard, Gold, Platinum, Enterprise); self-managed and Serverless priced separately
No Workplace Search-specific price; solution page only 'Contact sales'.
Elastic Cloud Hosted 'as low as' USD 99 (Standard), 114 (Gold), 131 (Platinum), 184 (Enterprise) per month for 120 GB storage / 2 zones base configuration
Unit: USD Platform price, not a per-user search licence. Tier required for document level security not named.
Customer selects deployment region; European Cloud Hosted regions: AWS Frankfurt, Zurich, Stockholm, Milan, Ireland, London, Paris; Azure Ireland, Netherlands, Paris, London; GCP Finland, Belgium, London, Frankfurt, Netherlands, Paris; or self-managed on-prem
Trust FAQ (https://www.elastic.co/trust/faq): Elastic Inference Service requests 'are routed to the nearest available region', which may differ from deployment region.
Not used for training
Quote: 'Elastic does not use your data to train its AI models.' Built-in models: 'Customer data is never used to train, validate, or fine-tune these models.' Customer-configured third-party LLMs governed by customer's agreement with that provider.
25 named connectors plus customised connector: Azure Blob Storage, Box, Confluence, Dropbox, GitHub, Gmail, Google Cloud Storage, Google Drive, Jira, Microsoft SQL, MongoDB, MySQL, Network drive, OneDrive, Oracle, Outlook, PostgreSQL, S3, Salesforce, ServiceNow, SharePoint Online, SharePoint Server, Slack, Teams, Zoom
Count derived from listed logos (26 entries incl. 'Customized connector'). Self-managed connectors and APIs for custom on-prem sources.
Listed: ISO/IEC 27001, 27017, 27018; SOC 2 and SOC 3 (attestations); CSA STAR; PCI DSS; TISAX; FedRAMP High/Moderate; Cyber Essentials Plus; IRAP; scope and auditor not named on trust page
Confirm Cloud Hosted versus Serverless certificate scope. HIPAA is a law.
Standalone Workplace Search deprecated in 9.0; connector document level security is beta ('provided as-is with no warranties', not under GA support SLA); Workplace Search uses its own separate permissions system
Deprecation from https://www.elastic.co/blog/search-for-enterprise-serverless.
Genesys Cloud AI Genesys · Product
A contact-centre platform with AI agents (“Agentic Virtual Agent”), copilots for agents and supervisors, and AI Studio.
Per user per month, bil… ● Genesys Cloud runs on A… ● Genesys Cloud AI
Genesys · Product
Facts, sources and conditions A contact-centre platform with AI agents (“Agentic Virtual Agent”), copilots for agents and supervisors, and AI Studio.
Where it sits on the map · 1 · 1 Contact centres that want the contact-centre platform and its AI features from one provider.
ETHORITY analysis
Per user per month, billed annually; certain AI features consume “AI Experience tokens”.
USD list prices per user per month, billed annually: CX 1 $75, CX 2 $115, CX 3 $155, CX 4 $240. CX 4 includes 30 AI Experience tokens per named agent. EUR prices not shown in the text read.
USD list prices.
Genesys Cloud runs on Amazon Web Services in 21 regions (Genesys, Feb 2026); an offering on the AWS European Sovereign Cloud was announced for mid-2026. Which region applies to your organisation is set by contract.
Only with opt-in
Quote: 'Customers can consent to participate in service improvements through a rigorously controlled process'; default opted out per MSA; data 'sampled and fully anonymized' before training. Third-party LLM training not addressed.
Messaging: Facebook Messenger, Instagram, X DMs, WhatsApp; CX Cloud from Genesys and Salesforce; embedded clients; customer-provided LLM for Agent Copilot summarization (planned 16 Mar 2026); STT/TTS via Microsoft, Google, Amazon Transcribe, Deepgram, ElevenLabs.
BYO LLM from https://help.genesys.cloud/announcements/genesys-cloud/agent-copilot-support-for-customer-provided-llms/.
Genesys lists 33 compliance items and classifies each. Certifications include ISO/IEC 27001:2022, 27017, 27018, 22301 and 42001:2023. Attestations include SOC 1, SOC 2 and SOC 3 Type 2, HITRUST and C5 (Genesys files C5 under “Certification”, but C5 is an audit attestation, not a BSI certification). PCI DSS and several government authorizations are listed; GDPR, HIPAA, DORA and the EU Data Act are listed as laws. Which items cover the AI features is not stated.
Platform-level scope; confirm AI-feature coverage. C5 is an attestation.
AI features consume tokens subject to Fair Use policy; screen recordings retained max 365 days; customer-LLM summaries limited to languages the chosen LLM supports; sentiment analysis in German marked with dialect-improvement asterisk.
Screen-recording limit from https://help.genesys.cloud/?p=360248; language notes from https://help.genesys.cloud/articles/genesys-cloud-supported-languages/. No accuracy limitations documented on pages read.
Grok Build (grok-code-fast-1) xAI · Model
xAI's grok-code-fast-1 (now aliased to grok-build-0.1) is a fast reasoning model with native tool calling, built to run multi-step agentic coding loops inside coding tools via an OpenAI-compatible API.
256,000 ● Conditions apply – open for details xAI API (OpenAI-compati… ● Conditions apply – open for details Hosted API (SaaS) only … ● Conditions apply – open for details Model card reports safe… ○ Conditions apply – open for details Grok Build (grok-code-fast-1)
xAI · Model
Facts, sources and conditions xAI's grok-code-fast-1 (now aliased to grok-build-0.1) is a fast reasoning model with native tool calling, built to run multi-step agentic coding loops inside coding tools via an OpenAI-compatible API.
The model page is now headed Grok Build0.1 (grok-build-0.1), with grok-code-fast-1 listed as an alias. Provider pages name SpaceXSI. https://docs.x.ai/developers/models/grok-code-fast-1
Where it sits on the map · 1 · 1 Agentic coding model (tool-calling code agent backend)
Derived from prompt-engineering guide and model card.
Per 1M tokens (input, cached input, output)
Unit: USD Requests above 200K context charged at different, unlisted rates; regional pricing mentioned but not shown.
Input $1.00 / cached input $0.20 / output $2.00 per 1M tokens
Unit: USD Grok Build 0.1, alias grok-code-fast-1, as described on the model page.
No EU region stated; model page lists regions us-east-1 and us-west-2. Default endpoint api.x.ai 'does not guarantee a processing region'; US-only endpoint (us.api.x.ai) serves only grok-4.7 and grok-4.6.
Region statements from https://docs.x.ai/developers/faq/security. No EU processing option found in reviewed pages.
Not used for training — "never trains on your API inputs or outputs without your explicit permission"
Enterprise ToS s.3.1: will not use any User Content to train foundation models/LLMs (https://x.ai/legal/terms-of-service-enterprise). Qualifier: 'without your explicit permission'.
OpenAI-compatible API with native tool/function calling; described as a pair-programmer 'inside most common coding tools' and offered via unnamed 'launch partners' / third-party partners. No IDE, Git host or CI/CD named on reviewed pages.
The model card describes availability through third-party partners.
SOC 2 Type 2 (self-stated: "We are SOC 2 Type 2 compliant"); scope and auditor not stated
Provider describes SOC 2 access under NDA. SOC 2 is an attestation.
Reasoning traces only accessible in streaming mode; XML-based tool-call output 'may hurt performance'; Grok 4 positioned as better for one-shot Q&A and deep debugging; Batch API not supported; model card reports MASK dishonesty rate 71.9 (higher than Grok 4) and agentic misuse risk.
Batch API from model page; MASK and risk statements from https://data.x.ai/2025-08-26-grok-code-fast-1-model-card.pdf.
Context window 256,000
Unit: tokens Pricing differs above 200K. Model card does not state context window.
Model access xAI API (OpenAI-compatible) and third-party partners; regions us-east-1, us-west-2; no cloud marketplace or download stated
Partners from model card https://data.x.ai/2025-08-26-grok-code-fast-1-model-card.pdf; not named.
Deployment requirements Hosted API (SaaS) only on reviewed pages; no open weights or self-hosting stated
No self-hosted/VPC option found in reviewed sources.
Published evaluations and scope Model card reports safety/dual-use evals only, e.g. "AgentHarm answer rate: 17.0", "AgentDojo attack success rate: 26.9", "MASK dishonesty rate: 71.9", "CyBench unguided success rate: 22.5". Guide claims "up to 4x the speed and 1/10th the cost of other leading agentic models". No coding benchmark (e.g. SWE-bench) on reviewed pages.
Performance and cost figures are provider claims; test against your own workload.
Documented limitations Reasoning traces only accessible in streaming mode; XML-based tool-call output 'may hurt performance'; Grok 4 positioned as better for one-shot Q&A and deep debugging; Batch API not supported; model card reports MASK dishonesty rate 71.9 (higher than Grok 4) and agentic misuse risk.
Batch API from model page; MASK and risk statements from https://data.x.ai/2025-08-26-grok-code-fast-1-model-card.pdf.
IN innoGPT innoGPT · Platform
innoGPT is a German multi-model AI workspace offering 50+ models (GPT, Claude, Gemini, Mistral) with automatic 'Smart Select' routing, AWS Frankfurt hosting and SharePoint/Drive knowledge search.
Per user per month ○ Conditions apply – open for details EU (AWS Frankfurt) ○ Sources disagree – open for details IN innoGPT
innoGPT · Platform
Facts, sources and conditions innoGPT is a German multi-model AI workspace offering 50+ models (GPT, Claude, Gemini, Mistral) with automatic 'Smart Select' routing, AWS Frankfurt hosting and SharePoint/Drive knowledge search.
The model list and SharePoint/Drive integrations are described on separate provider pages. https://www.innogpt.de/en/models https://www.innogpt.de/en/integrations.
Where it sits on the map · 1 · 1 Multi-model GenAI chat workspace with company knowledge connectors
Derived from homepage, pricing and integrations pages.
Per user per month (Business, Business Max, § 203 licence); Enterprise custom with paid pilot; add-ons per month, per generation or per token
Yearly billing gives one month free. Premium model usage weighted by model; top-ups EUR 5 to 1,000 one-time.
Business EUR 25/user/month; Business Max EUR 109/user/month; § 203 licence EUR 25/user/month (min. 20 licences, on request); Enterprise custom, 3-month pilot from EUR 8,500; PII filter add-on from EUR 69/month
Unit: EUR Prices excl. VAT. Page metadata also mentions a Personal plan from EUR 19, details not shown.
EU hosting; AWS Frankfurt named for data; '(most) AI models are hosted in Europe'; models page tags each model EU or 'Global'; § 203 plan lists 'EU hosting in Berlin'
Where the provider's own sources disagree The provider’s own sources say different things here. CONTRADICTION/ambiguity: pricing page § 203 plan says 'EU hosting in Berlin' while footer and security page name AWS Frankfurt. Models page (https://www.innogpt.de/en/models) tags several models 'Global' (e.g. gemini-3-pro-preview, sonar-pro, llama-4-maverick) without defining location. Trust centre (https://trust.innogpt.de) states no hosting region.
Not used for training
Quote: 'Neither we nor our technology partners (like Microsoft) ever use your inputs or the generated outputs to train AI models.' Pricing page: 'No data used for model training'; § 203 plan: 'Contractual guarantee: no model training on your data.'
'1,000+' integrations; named: Google Drive, SharePoint, OneDrive, Notion, Confluence, Slack, Microsoft Teams, Salesforce, HubSpot, Jira, Zapier, Linear, Asana, Airtable, GitHub, Monday; native MCP support; Custom Actions via OpenAPI spec; OpenAI-compatible API
Where the provider's own sources disagree The provider’s own sources say different things here. Count '1,000+' stated; integrations page also says '+ 1,000 more integrations available' beneath 16 logos (wording inconsistency). Mechanism behind the 1,000+ figure not described.
ISO/IEC 27001:2022 (stated as certified); scope and certification body not named
Homepage says 'Independently audited information security management' without naming auditor. Not third-party verified. GDPR listed alongside is a law, not a certification.
Only '(most)' AI models hosted in Europe; Custom Actions not available in free chat (assistants/agents only); Workflows marked 'soon'; trial limited to standard models and not § 203 compliant; one account per provider and user for integrations; SCIM provisioning 'in Arbeit'
Trial limits from https://www.innogpt.de/en/pricing; integration account limit from https://docs.innogpt.de/de/articles/5584558-integrationen; SCIM from https://docs.innogpt.de/de/articles/3370697-nutzerverwaltung-mitglieder-rollen-berechtigungen; '(most)' from security page.
Intercom Fin Intercom · Product
An AI agent that answers customer requests and is billed per outcome; Intercom states Fin works with “no seats required”.
Per Fin outcome. ● Intercom offers hosting… ● Intercom Fin
Intercom · Product
Facts, sources and conditions An AI agent that answers customer requests and is billed per outcome; Intercom states Fin works with “no seats required”.
Where it sits on the map · 1 · 1 Service teams that want result-based pricing and a published price per outcome.
ETHORITY analysis
Per Fin outcome.
From $0.99 per Fin outcome. Help-desk seats if needed: Essential $29, Advanced $85, Expert $132 per seat per month; the billing basis (annual or monthly) of these seat prices is to be confirmed — the page offers both.
USD starting price.
Intercom offers hosting regions US, EU and Australia (“US · EU · AU”), “in the region your residency needs require”.
Used unless you opt out
Fin CX Models trained on de-identified production Fin conversation logs; opt-out via Settings > Workspace > Security, processed within 30 days. Auto-excluded: BAA, EU/AU hosted workspaces, prior opt-outs, some negotiated contracts, trials. Third-party providers contractually barred from training (security page). Quote: 'used solely to improve model performance for our AI Products.'
Works with other helpdesks: 'Use Fin with your current helpdesk including Salesforce and more'; channels Messenger (web/iOS/Android), Email, WhatsApp, SMS, Facebook, Instagram; hands off to existing helpdesk.
Channels from https://intercom.com/help/en/articles/7837535-fin-faqs. Zendesk not named on pages read.
Intercom shows SOC 2 Type II, ISO 27001, ISO 27701 and ISO 42001 badges; Fin AI Agent is described as AIUC-1 certified. HIPAA support is listed for Expert. Confirm each report or certificate's product scope.
Cannot book/schedule meetings; cannot reliably calculate dates/business days (informational Fin); no access to macros; API conversations only via Workflow; cannot reassign from teammate inbox; unlisted languages 'unpredictable'; no feature to redact personal data before it reaches AI providers; customer responsible for testing before go-live.
Redaction/testing limits from https://www.intercom.com/help/en/articles/10213899-intercom-ai-products-a-guide-for-customer-legal-and-security-teams.
Logicc is a browser-based multi-model AI workspace (OpenAI, Anthropic, Google, Mistral) for regulated industries, storing chats and files on Hetzner servers in Germany and offering § 203 StGB agreements.
Per user/mo or annual ● Conditions apply – open for details DE storage · EU models ○ Conditions apply – open for details Facts, sources and conditions Logicc is a browser-based multi-model AI workspace (OpenAI, Anthropic, Google, Mistral) for regulated industries, storing chats and files on Hetzner servers in Germany and offering § 203 StGB agreements.
Model availability and the agreement concerning §203 StGB are described on separate provider pages; confirm the applicable contractual scope. https://www.logicc.com/en/platform/models https://www.logicc.com/en/pricing-plans.
Where it sits on the map · 1 · 1 Secure multi-model AI chat and document analysis for teams
Derived from platform and pricing pages.
Per user per month (annual or monthly billing); Enterprise custom
Fair monthly usage budget per organisation that grows with team size. Monthly term, auto-renews.
Business EUR 29.90 (EUR 23.92 annual); Secure+ EUR 49.90 (EUR 39.92 annual); Max EUR 89.90 (EUR 71.92 annual) per user/month excl. VAT; Enterprise custom
Unit: EUR Premium Integrations add-on 'Price to be announced' (https://www.logicc.com/en/platform/premium-integrations).
Chats, documents and account data stored exclusively in Germany (Hetzner, Nuremberg and Falkenstein); model processing: OpenAI on Azure Sweden, Claude/Gemini on Google Cloud Germany and Belgium, Mistral 'EU'
Security page: 'Storage exclusively in Germany', 'Deliberately no US host'.
Not used for training
Quote: 'Your data doesn't train any AI model, ruled out both contractually and technically.' Zero Data Retention at model providers 'as far as technically possible'.
'Connect existing tools and data sources' with logos incl. Microsoft, Google, Slack, Notion, Dropbox, HubSpot, Salesforce, Asana, Trello, Zapier, Loom; Model API; custom API integrations (Enterprise); 5 industry Premium Integrations (Gesetze im Internet, commentaries/case law, medical billing, medical research, DATEV) all 'In preparation'
Connector scope per logo not described; no count of general connectors stated. Premium list from https://www.logicc.com/en/platform/premium-integrations.
No own certification stated. Hosting partner's data centres ISO 27001 certified (issuer not named); hosting partner's cloud infrastructure attested under BSI C5:2020 Type 2
Both relate to the cloud host (Hetzner), not Logicc itself. Certificates and audit reports available from Secure+ plan. Not third-party verified.
Up to 1,000 pages per document; desktop and mobile apps 'in preparation' (browser only); top models (GPT-5.6 Sol, Claude Opus) Max plan only; some model providers not covered under § 203 StGB agreement; fair-use budget per plan
Page-limit and app status from https://www.logicc.com/en/platform/chat.
PR Lumo for Business Proton · Product
Proton's Lumo for Business is a team AI chat assistant running open-source models (Qwen, GLM, Apertus) on Proton-controlled European servers, with no server-side chat logs and zero-access-encrypted history.
Per user per month ○ Conditions apply – open for details Europe ('Lumo is hosted… ○ Conditions apply – open for details PR Lumo for Business
Proton · Product
Facts, sources and conditions Proton's Lumo for Business is a team AI chat assistant running open-source models (Qwen, GLM, Apertus) on Proton-controlled European servers, with no server-side chat logs and zero-access-encrypted history.
The privacy guide names models and Proton-controlled servers. The Lumo Business article states European hosting; confirm the applicable product and plan. https://proton.me/support/lumo-privacy https://proton.me/business/blog/lumo-business.
Where it sits on the map · 1 · 1 Confidential AI chat assistant for teams
Derived from product page and FAQ.
Per user per month (billed yearly or monthly); CHF, EUR, USD
Blog: 'affordable for teams of any size with no minimums'.
Lumo Professional USD 11.99/user/month billed yearly (USD 143.88/year; list USD 14.99); Workspace Premium (bundle with Mail, Drive, Pass, VPN etc.) USD 19.99/user/month billed yearly (USD 239.88/year; list USD 24.99)
Unit: USD Taxes may apply. Confirm the monthly price and the EUR/CHF currency selection in your quote.
Europe ('Lumo is hosted entirely in Europe'); models run 'exclusively on servers Proton controls'; specific countries/data centres not named
Lumo privacy policy (https://proton.me/lumo/privacy-policy, modified Sept 17 2026) does not state Proton's own server locations; third-party processors (Zendesk, Stripe, Chargebee, PayPal, Atlassian) are in US/EU/UK/Singapore for support/payment data.
Only with opt-in
Where the provider's own sources disagree The provider’s own sources say different things here. Quote: 'never keeps your data or uses it to train LLMs, except with your express permission' (opt-in anonymised Apertus feedback shared with ETH Zurich/EPFL). CONTRADICTION in wording: https://proton.me/business/lumo says 'Lumo never trains AI models on your chats' without the opt-in exception.
Proton Drive (add PDFs or Docs to conversations); optional web search via unnamed partner search APIs; used alongside Proton Mail, Calendar, Drive, Pass, VPN, Meet
No third-party business-system connectors (e.g. SharePoint, Google Drive) named. Blog: 'There's nothing you need to integrate.'
ISO 27001 certified and SOC 2 Type II audited (attestation), stated for 'All Proton software'; issuer/auditor, dates and scope details not named
Lumo page references 'Proton's ISO 27001 and SOC 2 certifications'. Not third-party verified. GDPR/HIPAA are laws, not certifications.
Web search optional and off by default; usage limits described only as 'More'/'Unlimited' without figures; chat languages limited to 11 listed
No explicit provider-stated accuracy or functional limitations found on product, pricing, privacy support page, blog or FAQ.
MeinGPT is a Hetzner-hosted German AI platform for Mittelstand firms offering multi-model chat and assistants, plus an 'Outpost' that connects on-premise systems through an outbound-only connection.
Per user/mo + usage ● Conditions apply – open for details Mainly Germany · EU ● Conditions apply – open for details SM meinGPT
SelectCode (MeinGPT) · Platform
Facts, sources and conditions MeinGPT is a Hetzner-hosted German AI platform for Mittelstand firms offering multi-model chat and assistants, plus an 'Outpost' that connects on-premise systems through an outbound-only connection.
Where it sits on the map · 1 · 1 Company-wide multi-model AI chat, assistants and knowledge connectors for SMEs
Derived from home and pricing pages.
Platform licence per user per month in progressive tiers, plus separately billed model budget (credits/usage-based per 1M tokens) and add-ons
Usage prices per model in https://meingpt.com/en/docs/admin-guide/pricing. 1 credit = EUR 0.01.
Licence per user/month (incl. 10% annual discount): 1-50 users EUR 18; 51-100 EUR 13.50; 101-250 EUR 9.90; 251-500 EUR 7.20; 501+ EUR 5.40. Model budget EUR 9/user fixed or shared pool (~EUR 2.31/user). Governance add-on EUR 1.35/user/month; Outpost from EUR 225/month
Unit: EUR Where the provider's own sources disagree The provider’s own sources say different things here. CONTRADICTION on same page: FAQ says first 50 users cost '20 euros each' vs EUR 18 in table (likely monthly vs annual). VAT treatment not stated. Model token prices e.g. GPT-5 EUR 1.50/12.00 per 1M in/out (docs pricing page).
Platform and data hosted primarily in Germany (Hetzner); 'Hosted in Germany & the EU'; with DPA 2.0 all AI models run in the EU, otherwise non-EU models flagged in model picker and only used if admin selects them
Security page: 'Hetzner data centers in Germany'. Docs pricing page: GPT-6 Astra 'runs at OpenAI with EU data residency'.
Not used for training
Docs FAQ states customer data is not used for AI training; Zero Data Retention at model providers. Security page (https://meingpt.com/en/security) poses the training FAQ question without visible answer.
8 integration entry points in docs: Microsoft 365 (SharePoint, OneDrive, Outlook, Teams), Google Workspace (Drive, Gmail, Calendar), Atlassian (Confluence), file shares (WebDAV; SMB via Outpost), SQL/PostgreSQL via MCP, business SaaS (HubSpot, Salesforce, Stripe), web crawler, custom MCP/internal APIs; also Amazon S3, Make, n8n, Power Automate
Home page additionally lists SAP, Oracle, Snowflake, Jira, AWS, Zoom; Outpost mentions DATEV, DMS.
ISO/IEC 27001:2022-certified ISMS claimed; issuer, scope and certificate not shown. Data centres (Hetzner) ISO 27001 certified
Where the provider's own sources disagree The provider’s own sources say different things here. Provider pages differ between ISO alignment, certification and data-centre certification; confirm the certified entity and scope.
SharePoint direct connector read-only, no guaranteed subfolder restriction; SharePoint Lists not synced; SharePoint data-pool source not available for on-prem Outpost; Cloud Vault index not live (15-min refresh); SMB/local files need Enterprise Outpost; some features beta/coming soon
Also from https://meingpt.com/en/integrationen/sharepoint-ki and https://meingpt.com/en/docs/integrations.
MI Mindverse Studio Mindverse · Platform
Mindverse Studio is a Berlin-built, credit-metered browser AI workspace hosted in Germany, combining EU-endpoint OpenAI/Anthropic/Google models and German-hosted open models with knowledge bases and 500+ Composio integrations.
Per user/mo + credits ○ Conditions apply – open for details Hosting: Germany ○ Conditions apply – open for details MI Mindverse Studio
Mindverse · Platform
Facts, sources and conditions Mindverse Studio is a Berlin-built, credit-metered browser AI workspace hosted in Germany, combining EU-endpoint OpenAI/Anthropic/Google models and German-hosted open models with knowledge bases and 500+ Composio integrations.
The pricing page describes credits; the privacy policy names Relativity GmbH, Berlin, as operator. https://www.mind-verse.de/en/preise
Where it sits on the map · 1 · 1 Multi-model GenAI workspace with assistants, knowledge bases and agents
Derived from product and pricing pages.
Per user per month with monthly credit allowance; extra credit packs; Enterprise on quote
Credits consumed per chat/image/video by model tier (Standard from 3 credits, Expert from 10, video 64 credits/second).
Starter EUR 19 (EUR 15.20 annual)/user/month; Pro EUR 59 (EUR 47.20 annual)/user/month; Team EUR 49 (EUR 39.20 annual)/user/month from 3 users; Enterprise on request; credit packs EUR 10/1,050, EUR 50/5,750, EUR 200/24,000
Unit: EUR Excl. VAT. Monthly credits: Starter 200, Pro 800, Team 1,000 per user.
Germany: 'Hosting: 100% Germany' / 'hosted exclusively in Germany'; sovereign models on German infrastructure; OpenAI/Anthropic/Google via 'EU / Germany' endpoints
Website privacy policy (https://www.mind-verse.de/legals/datenschutz) names AWS with location Deutschland for web hosting/backend. Cloud platform behind EU model endpoints not named.
Not used for training
Quote: 'Your content, documents, and prompts are never used to train LLMs.' FAQ: 'Your prompts, documents and knowledge bases are never used to train AI models.'
'Over 500 integrations' via Composio (CRM to project management); custom connections to Sitecore, DMS, CRM and line-of-business systems (Enterprise)
Pricing describes 500+ integrations; confirm the named connectors required for your workflow.
None held: ISMS 'built to ISO 27001:2022' but 'not currently certified by an accredited body'; 'SOC 2-oriented processes' 'without an attested SOC 2 report'; data centres in Germany stated as certified (standard/issuer not named)
Hosting data-centre certification would be the hosting provider's, not Mindverse's own. Hamburg Declaration on Responsible AI is a commitment, not a certification.
Expert and Sovereign models not available on Starter; usage capped by monthly credits; inference on German infrastructure 'on request'
No functional/accuracy limitations stated on FAQ, product or security pages.
