In plain language.
Describe the issue in your own words — a plain question or a pleaded point. Select the jurisdiction, or let the engine detect it.
Ask any legal question across five common law jurisdictions. Get structured analysis with verified case citations, commercial assessment, and practical next steps — typically in about three minutes, not hours.
UK judgment text stored for pinpoints: 0 / — (a pinpoint is emitted only from a stored paragraph hash).
Specimen-cited authorities: Gate T — pending and terminal counted separately. A pinpoint is emitted only from a stored paragraph hash.
The engine types, flags, answers, cites and scores a real enquiry — the same sequence every user sees.
Strong grounds for a worldwide freezing order. On a good arguable case plus a real risk of dissipation — and the post-award transfers are direct evidence of that risk — the court can restrain the respondent’s assets globally. The English order binds the respondent personally but does not by itself attach assets in Singapore or Hong Kong: local recognition or enforcement steps are required there, and the order will carry the standard provisos limiting its effect abroad until local leave is obtained.
Commercial & tactical: the without-notice route imposes a strict duty of full and frank disclosure — any material omission is the most likely path to discharge, with costs. Budget for a cross-undertaking in damages and expect the court to require fortification given the cross-border footprint. Net of foreign enforcement costs, recovery remains viable against the £6.2m; sequencing the Singapore and Hong Kong steps early is critical to prevent further dissipation.
The post-award transfers through the BVI holding company are themselves the clearest evidence of dissipation risk — the respondent’s own evasive conduct supplies the ground for the order.
Describe the issue in your own words — a plain question or a pleaded point. Select the jurisdiction, or let the engine detect it.
The corpus holds 53,100+ authorities across five jurisdictions. Chat and analysis retrieve from the 19,300+ that have been confirmed against a primary source; remaining holdings stay quarantined pending confirmation, and the verified set grows as sweeps clear them. Citations are checked for existence before you see them; verification status is shown because no automated check is infallible.
Commercial assessment, tactical next steps, the opponent’s weakest point — everything you need to walk into court or the settlement table.
Federal, state and territory coverage where applicable. The same verification pipeline runs across every jurisdiction we support.
A corpus of 53,100+ authorities spanning landmark cases and modern appellate decisions across all five jurisdictions. 19,300+ of those are confirmed against a primary source and used in chat retrieval; remaining holdings stay quarantined pending the same confirmation, and the verified set grows as sweeps clear them. Curated civil-justice landmarks deepen coverage of leading procedural doctrine; family and matrimonial authorities are held on the same terms (US domestic relations is state law).
Citations in every answer are checked for existence against the corpus and, where source policy permits, live public endpoints. Verified citations get a ✓ badge; anything unconfirmed is flagged; failures are stripped. That confirms existence — not good law or proposition accuracy.
Every response with financial figures includes a costs estimate, net recovery calculation, and viability assessment. The question every client actually asks — answered automatically.
“The other side will argue…” is included in every analytical response. Know your weaknesses before the other party finds them.
Upload pleadings for line-by-line critique. Structural feedback, missing issues, and tactical suggestions. Advocate and Chambers tiers.
Statutory demands, limitation deadlines, and asset dissipation risks are automatically flagged before the analysis begins, so time-critical issues surface first.
CommonBench compared with a generic AI chat and a junior lawyer briefed for four hours.
| Capability | Generic AI Chat | CommonBench | Junior Lawyer (4 hrs) |
|---|---|---|---|
| Citations verified against legal databases | — | ✓ | ✓ |
| Coverage across five common-law jurisdictions | — | ✓ | — |
| Chinese answers (Hong Kong & Singapore only) | — | ✓ | ✓ |
| Commercial assessment (costs, net recovery) | — | ✓ | ✓ |
| Opponent’s weakest point identified | — | ✓ | ✓ |
| Urgency / limitation deadline flagged | — | ✓ | ✓ |
| Full analysis typically ready in about three minutes | ✓ | ✓ | — |
| Cost per enquiry | $0 | ~$0.40 | ~$1,000 |
For a single live matter, end to end.
For practitioners handling several matters at once.
Per-seat billing for firms. Hard isolation.
No. CommonBench provides AI-generated legal information for reference purposes only. It does not constitute legal advice and no solicitor-client relationship is created. Always verify with a qualified practitioner before acting on any output.
Every citation is checked for existence against our corpus and, where source policy permits, against live public endpoints. Verified citations display a ✓ badge; unconfirmed citations are flagged; failures are stripped before display. Existence checks do not prove good law or that a proposition is correct. No automated check is infallible — confirm any authority before relying on it in proceedings.
England & Wales, the United States (Supreme Court and federal appellate, with selective state coverage), Hong Kong, Singapore, and Australia (federal and state appellate courts). Scotland and Northern Ireland queries may be accepted under the UK selector, but the corpus and procedural tooling are England-and-Wales-centric — not full Scots or Northern Irish coverage. Curated civil-procedure and family/matrimonial doctrine is included; it is not a substitute for the current White Book, local forms, or specialist family practice. US state-specific research — including family law, which is state law — should be verified locally.
Yes — for Hong Kong and Singapore only. When you select Hong Kong or Singapore, you can write in Traditional Chinese (繁體中文) or Simplified Chinese (简体中文) and receive a legal analysis in the matching script. For England & Wales, Australia, and the United States, answers are in English even if the question is asked in Chinese: those systems are not treated as Chinese advice languages here. Chinese answers use Hong Kong / Singapore common-law register; they are not Mainland PRC civil-procedure advice.
Every tier runs the same engine and the same verified corpus. Completeness does not change with price — every tier must surface the material issues, the adverse points, and the next steps. What you pay for is professional judgment (junior-associate on Starter, partner on Advocate, senior-counsel forensic synthesis on Chambers) plus capacity and workflow. Starter covers one live dispute (~250 questions/month). Advocate adds document upload, costs & ADR sequencing, priority processing (~600/month). Chambers is per-seat billing for firms with hard isolation (~1,200/seat/month).
A full reasoned analysis typically takes about three minutes. Deeper multi-issue questions can take longer (often two to six minutes) because the model reasons before it writes; shorter starter queries are usually faster. Citation checks and self-critique finish after the answer starts streaming, so you see the substance as soon as the model finishes drafting.
Plans will be month-to-month with no lock-in. Access is not open yet. Leave an email for a cohort; once you have a subscription you can cancel from your account and keep access until the end of that billing period.
Access is not open yet. Leave an email for a cohort — we open by jurisdiction.
Access is not open yet. We’ll be in touch as we open by jurisdiction. See pricing for what it will cost.