Counsel+ is included with paid subscriptions and is designed to improve correctness and consistency across Swiss legal research, drafting, and document review tasks. It is intended to help legal professionals review outputs more efficiently within their existing workflows.

Internal evaluation indicates output at approximately a 5.8 level on the Swiss grading scale.

🔒 Included with all paid subscriptions.

Counsel+ is included with paid subscriptions and is designed to improve correctness and consistency across Swiss legal research, drafting, and document review tasks. It is intended to help legal professionals review outputs more efficiently within their existing workflows.

Internal evaluation indicates output at approximately a 5.8 level on the Swiss grading scale.

🔒 Unavailable for your account.

Claude is a premium model with substantially higher running costs, so it's available exclusively on the Professional (CHF 200) plan.

Our premium model, included with the Professional plan. Built for complex, multi-layered questions. Served exclusively from the EU region with Zero Data Retention (ZDR).

Where the default model infers what you want and adjusts its writing length and style, Claude follows instructions more literally. If you do not specify how you want the answer to be written, it will overthink and cover all options, including adjacent issues you never raised.

Specify the desired writing style:

• Form: memo, short opinion, checklist, or the answer first and the reasons after.
• Length: one page, three paragraphs, or as detailed as possible.
• Scope: what to include, and what to leave out.
• Standpoint: whose side you are on, and whether you need a cautious or assertive stance.

All AI models make mistakes. Always check the answer and its sources before relying on it.

Anthropic released Claude Opus 5.5 yesterday, and we're already bringing it to your Legislator workspace.

That means another powerful model to choose from when working on exceptionally complex tasks that require deeper, multi-step reasoning. However, it is much slower than our default model, and its writing style is worse. In general, we recommend using the default model.

With multiple leading AI models available in one place, you can choose a specific model from the model selector, or let Legislator select the model best suited to your task, so you can focus on the work rather than which model to use.

Opus 5.5 is served from EU by our partner provider using Zero Data Retention.

Served from the EU with Zero Data Retention (ZDR). The model provider does not keep your prompts or answers.

All AI models make mistakes. Always check the answer and its sources before relying on it.

Preliminary legal assessment (~1 min).

Matches your facts against decisions across all 26 cantonal courts, identifying cited Federal Supreme Court (BGE) rulings, federal statutes (SR), and recent developments. Ideal for initial case triage before requesting in-depth research.

All AI models make mistakes. Always check the answer and its sources before relying on it.

The original default mode (~5–10 min).

Comprehensive examination across federal and cantonal legislation, 350 Swiss legal commentaries (Helbing Lichtenhahn, Dike), and Federal Supreme Court jurisprudence, producing a structured legal opinion. Recommended for routine inquiries.

All AI models make mistakes. Always check the answer and its sources before relying on it.

Dual-perspective research (~10 min).

Conducts two full investigations: first establishing the legal analysis for your position, then analyzing the case from the opposing party’s perspective to uncover weaknesses, vulnerabilities, and risks. Both viewpoints are synthesized into a single assessment.

All AI models make mistakes. Always check the answer and its sources before relying on it.

Exhaustive dual-perspective investigation (~10–15 min).

Runs our deepest dual-perspective analysis. Investigates edge cases, minority views, and alternative remedies from both sides, subjecting your case to exhaustive examination from the opposing party's viewpoint to uncover every potential exposure.

All AI models make mistakes. Always check the answer and its sources before relying on it.

Direct single-perspective research (~5 min).

Focuses solely on researching your position. Skips the opposing-party weakness analysis and the upfront precedent and commentary scan to deliver a faster answer.

All AI models make mistakes. Always check the answer and its sources before relying on it.

Draft legal memorandum (~10 min).

Drafts a formal legal memorandum with statement of facts, formulation of legal questions, substantive discussion, and full citation apparatus, rather than a conversational reply.

All AI models make mistakes. Always check the answer and its sources before relying on it.

This option is not needed for provider privacy: in all modes, we maintain Zero Data Retention (ZDR) agreements with all model providers and use EU endpoints. Providers do not store or train on user data.

With this option, data is completely erased from our servers after 24 hours, including backups, caches, and logs. This makes recovery technically impossible in the event of a subpoena, court order, or government request.