As of 28 Oct 2026, the panel rates Claude for Work 3/5 and Langdock 4/5 in AI Agents & Assistants. Weighted by what each judge cares about, 0 of the 6 judges pick Claude for Work and 6 pick Langdock. Claude for Work scores higher on 1 of 7 criteria (Admin, identity & access controls). Langdock scores higher on 6 of 7 criteria (Models & processing location, European sovereignty, Agent capabilities & guardrails, plus 3 more). Provenance — Claude for Work: Rest of world; Langdock: EU-Made. Ratings are the AI panel’s opinion, formed from public evidence: how scores are made.
Langdock holds the wider margins: processing and models 8.2 to 3.7, sovereignty 8.0 to 3.8, agent capabilities 8.0 to 7.0, pricing transparency 7.2 to 6.2, integrations and knowledge 7.0 to 6.7 — six judges lean Langdock on the first three, five on pricing transparency, two on integrations and knowledge. Claude for Work is ahead on admin and access, 7.7 to 7.2, three judges leaning its way, three tying. Data use and training is the only criterion with judges leaning both ways: 7.8 to 8.0, one for Claude for Work, two for Langdock, three ties. Published weighted totals favour Langdock on the five judges with numbers shown, 7.5–7.9 against 5.6–6.8; the customer-service director's row shows none. Sovereignty attributes as they stand: Claude for Work — US legal entity, US-default data residency; Langdock — German legal entity, EU-only data residency; ownership unknown for both.
Choose Claude for Work if
You need enterprise administration depth — single sign-on with domain capture, SCIM provisioning, custom roles with per-connector permission levels — where the table has Claude for Work ahead 7.7 to 7.2 with three judges leaning its way.
Your compliance process requires retention controls with every change tracked in audit logs; the verdict cites these for Claude for Work, while we found no public information on exportable audit logs for Langdock.
Your deciding criterion is data use and training, where the gap is the table's narrowest (7.8 to 8.0) and the bench splits one judge for Claude for Work, two for Langdock, three ties.
You must contract with a US legal entity and accept US-default data residency; the sovereignty attributes list Claude for Work as US jurisdiction with US-default residency.
Choose Langdock if
You need EU-only data residency and a German legal entity; the sovereignty attributes list Langdock as German jurisdiction with EU-only residency, and sovereignty runs 8.0 to 3.8 with six judges leaning Langdock.
Your workloads require EU-only inference by default, data at rest in EU data centres even for global models, and bring-your-own-key or own-cloud or on-premises deployment — processing and models stands at 8.2 to 3.7.
You need human confirmation on agent action nodes; agent capabilities runs 8.0 to 7.0 with six judges leaning Langdock, while Claude for Work's own pages state there is no approval step before a run starts.
Your data protection officer must sign off; their weighted total reads 7.7 for Langdock against 5.6 for Claude for Work.
Procurement needs published pricing clarity; pricing transparency runs 7.2 to 6.2 with five judges leaning Langdock and one tie.
Read this comparison as one judge. Each weighs the same scores by what they care about.
The Head of IT
Has to roll an AI assistant out to eight hundred staff without a second shadow-IT wave. Wants SSO and SCIM, groups that decide who gets which model, agent and connector, answers that respect SharePoint permissions, audit logs for the SIEM, and a deployment option that is not "trust us".
Claude for Work
Langdock
This judge's pick
Criterion by criterion
Data use, training & retention
Claude for Work
The commercial terms state that Anthropic may not train models on Customer Content from the Services, an opt-out is shown on all plans, and accepting the Commercial Terms also accepts the DPA. Enterprise admins can set custom retention for conversations, projects and artifacts from a 30-day minimum, data past the period is permanently deleted, every retention action is tracked in audit logs, abuse-monitoring retention for safety-flagged conversations is described openly, and SOC 2 and ISO 27001 are listed for the Enterprise plan. I hold it just under the top level because retention defaults to indefinite, the controls are Enterprise-only, and they exclude Claude Managed Agents and features running Claude Code in the cloud — exactly what an agent rollout would lean on.
Langdock
No-training is contractual, not just marketing: the terms state Langdock engages model providers only on agreements where those providers are "also prohibited from using Customer Content to develop, train or improve their AI Models", the DPA is auto-concluded, admins set a retention period with automatic chat deletion, per-artifact deletion behavior is documented, data is deleted 30 days after termination, and ISO 27001 plus SOC 2 Type II certifications are published. Zero-data-retention abuse monitoring is documented for the Azure-hosted models specifically; we found no public information on abuse-monitoring terms for the other model providers.
Models & processing location
Claude for Work
The model lineup is named — Opus, Sonnet, Haiku and Mythos, with Opus on paid plans only — and the privacy policy states that personal data is transferred to servers in the US or other countries outside the EEA and the UK. Claude is listed as available through AWS, Google Cloud and Microsoft Foundry, but the only regional pointer, to the Amazon Bedrock user guide, lists no specific regions on the captured page. We found no public information on an EU processing option, admin-selectable residency, bring-your-own-key, or an own-cloud or on-premises deployment for these plans.
Langdock
This is the model story I want for a regulated rollout: model-agnostic access to 40+ models, EU-only processing by default with global inference only when a customer actively selects such models, third-country transfer only on controller instruction written into the DPA, and bring-your-own-key plus dedicated own-cloud and on-premises deployments. A published per-model inference-region list and open-weight self-hosted models we found no public information on.
Admin, identity & access controls
Claude for Work
This covers most of what an eight-hundred-seat rollout needs: SAML single sign-on with domain capture, SCIM and just-in-time provisioning, and group-based custom roles that decide which models a role may use — with an effort-level cap and default model — which connectors it may reach at per-connector levels from Blocked to Always allow, and which admin areas from Billing to Privacy it manages, with groups synced from the identity provider via SCIM. Audit logs, usage analytics, spend controls with per-group monthly limits, and a Compliance API are listed for the Enterprise plan. We found no public information on data-loss-prevention policies, streaming audit logs to a SIEM, or pseudonymised reporting for works-council agreements, and permission-respecting retrieval is evidenced only for the Pipedrive CRM connector, not for SharePoint.
Langdock
SAML 2.0 with any identity provider, signed requests and domain verification, SCIM named on the Business plan, role-based access control with custom roles and group grants, IP restrictions, usage analytics togglable per workspace with CSV and API export, a governance console to review, approve or disable agents with compliance checks on every new version, and mirrored permissions that reflect source-tool access 1:1 — which answers my SharePoint concern directly. We found no public information on an exportable or SIEM-streamable audit log, which my SOC will want before eight hundred seats.
Integrations, connectors & knowledge
Claude for Work
Connectors for Gmail, Google Drive, Slack and Microsoft 365 are named and available on all plans, remote MCP is supported with a marketplace — the official Pipedrive MCP connector ships seventy-nine read-and-write CRM tools over OAuth and respects existing user permissions — and the Claude API, plugins and webhooks are documented. We found no public information on source citations for grounded answers, on sync speed or completeness, or on ticketing and ERP connectors.
Langdock
Open protocols are first-class: a directory of 79 verified MCP servers, custom MCP servers with several OAuth types, discovered tools converted into actions with confirmation flows, an MCP Apps SDK across major frameworks, and A2A agent communication, alongside a unified API, Outlook and Excel add-ins, Zapier reach, and agents usable in Slack and Teams. We found no public information on incremental connector sync or source citations on grounded answers.
Agent capabilities & guardrails
Claude for Work
The managed-agents documentation is unusually candid about mechanics and limits: workflows run in the background with session event streams, fan out to up to 1,000 agents and 64 concurrent threads per run, default to 24-hour lifetimes, pause on budget, usually retry a failed agent once, and connector tool calls can sit behind per-role approval levels with Needs approval as the default for new roles. The same pages state plainly that there is no approval step before a run starts. We found no public information on evaluation or test tooling, agent versioning, staged rollout, or per-agent quality and cost monitoring.
Langdock
Agents act through integration actions and workflows, with human confirmation on action nodes, approve/reject and form input nodes that pause runs, scheduling outside the builder, publish-and-draft versioning with a build-and-test mode, and per-agent instructions, analytics, costs and history visible in governance with automated checks on every published version. Fair-use limits and a stated no-warranty position on AI outputs are documented; evaluation against defined test sets before release we found no public information on.
European sovereignty
Claude for Work
European customers contract with and are controlled by Anthropic Ireland Limited, but the parent is the US company Anthropic PBC, the models behind the service are the vendor's own, and the privacy policy states personal data is transferred to servers in the US or other countries outside the EEA and the UK, with standard contractual clauses carrying the transfers. A subprocessor list is published by pointer. We found no public information on EU hosting or EU inference.
Langdock
Langdock GmbH registered in Berlin, exclusive jurisdiction Berlin, hosting in Frankfurt on Microsoft Azure with the EU Data Boundary, a published subprocessor list naming AWS, Azure, Google Cloud and OpenAI Ireland with locations, Sentry for error tracking under Data Privacy Framework and standard contractual clauses, EU-only inference by default with third-country transfer only on customer instruction, and an auto-concluded DPA with export and 30-day deletion documented. We found no public information on the ownership or parent-company situation, which keeps this short of the top level.
Pricing transparency
Claude for Work
Enterprise is published at $20 per seat per month plus usage billed at API rates, with a 20-seat minimum and annual commitment, and the tiers below are public with their inclusions — Pro at $17 per month with annual subscription discount ($200 billed up front) or $20 if billed monthly, Max from $100 per month. But usage limits are described only relatively, the usage rates sit on a separate Model Pricing Page the captured pages reference rather than show, and we found no public information on a Team seat price, so the real annual cost for eight hundred seats cannot be computed from the captured pages.
Langdock
For my seat count I can get most of the way there: "€23.20 Standard seat" and "€99.00 Business Max seat" per user per month excluding VAT with "5x more usage than Standard seat", governance at "€ 2.80 Per user / month (excl. VAT)" free until January 1, 2027, workflows from an included 2,500 runs to "€ 449.00 Per workspace / month (excl. VAT)" with 40,000 runs, per-token API prices, bring-your-own-key where you "pay Langdock only for the platform license", a 20% annual saving, and contract terms with termination notices and pro-rata mid-term additions. Enterprise is "Custom. Contact us", exact fair-use figures live in a separate policy, and we found no public information on overage rates, so the full annual cost is not computable from the pricing pages alone.
Sovereignty, side by side
Dimension
Claude for Work
Langdock
Legal entity
Not determined
Not determined
Ownership
Not determined
Not determined
Data residency
Not determined
Not determined
Subprocessors
Not determined
Not determined
Facts, side by side
Only facts both products carry under the same definition — anything else would not be a fair row.
As of 28 Oct 2026, the panel rates Langdock higher: 4/5 against 3/5. Weighted by what each judge cares about, 0 of the 6 judges pick Claude for Work and 6 pick Langdock. Which one fits depends on the criteria that matter to you; the table on this page shows both side by side, and each judge’s reading has its own page. Both were scored by the same AI bench under the same rubric, from public evidence only, as the methodology describes.
Is Claude for Work or Langdock European?
Claude for Work is not European by provenance (Rest of world). Langdock is European (EU-Made). Provenance is read from each company’s own legal pages: who builds it and who owns it.
Where do Claude for Work and Langdock store customer data?
Per the vendors’ own pages as captured — Claude for Work: its pages as captured do not say. Langdock: Default not stated; pages mention Europe.
Do Claude for Work and Langdock have a free plan?
As of 28 Oct 2026, Claude for Work’s own pages state a free plan; Langdock’s pages as captured do not. Each product page links the pricing page it was read from.