Tell us what happened and upload your medical records (multiple PDFs welcome). The AI Council — independent AI models, all running on AWS Bedrock under a signed HIPAA BAA, debating your case — reads every page and gives you a plain-language verdict on whether the medicine supports a claim. Fully automated: pay, the council debates, your AI Opinion & Verdict PDF lands in your inbox — usually within the hour, for a fraction of a lawyer's consult fee.
You don't need to understand your medical records. We turn them into a straight answer you can act on.
Answer a few simple questions and upload your medical records — any size, any format. Your information is kept private and confidential.
Clinical AI indexes every page, builds a timeline of your care, and flags what matters to your situation.
Using a family doctor's medical framework, the AI weighs your records against what typically supports a claim and produces a structured assessment — 100% automated, on AWS Bedrock under a signed HIPAA BAA, and never stored.
A plain-language report: do you likely have a case, how strong it is, and exactly what to do next — usually in a few days.
Stop paying for a lawyer's hour just to be told "we'd need a doctor to look at it." Start with the medicine — read by AI, on a family doctor's framework.
A clear go / maybe / no read on whether the medicine in your records supports a claim — and why.
includedWhat your records actually show, in words you understand — AI-generated on a family doctor's medical framework.
includedWhether it's worth seeing a lawyer, what kind, and what to bring — so you don't waste a paid consult.
includedA lawyer tells you the law. The medicine in your records is what actually decides most cases — and that's exactly what CaseCheck's AI reads, on a family doctor's framework.
| What you want to know | CaseCheck | Asking a lawyer first | A free AI chatbot |
|---|---|---|---|
| Does the medicine support a claim? | ✓ AI assessment | ✕ not medical | ✕ no structure |
| Built on a family doctor's medical framework | ✓ | ✕ | ✕ generic |
| Cost to find out | from $49 | $300–500+/hr | free |
| Answer you can actually trust | ✓ | ~ partial | ✕ |
| Turnaround | days | weeks for an appt | instant but unreliable |
More AIs, deeper debate. Every tier includes the full AI-to-AI discussion transcript and one final plain-language document with the relevant points in your case. Pay once — no subscription.
These are the actual models on the bench — with their real published scores. Click a model to toggle it and see its full scorecard with the source; watch the price: 5 selected → $49
| Who | Medicine — MedQA (USMLE-style) | Law — LegalBench (legal reasoning) |
|---|---|---|
| Expert physicians, same questions human reference | 87% Liévin et al. ↗ | |
| Human passing thresholds human reference — USMLE-style pass · bar exam MBE pass (different tests, shown for context) | ≈60% PLOS Digit. Health ↗ | ≈60% NCBE ↗ |
| Claude Opus 4.1 Anthropic · chairs the consensus | 92.5% vals.ai ↗ | 83.5% vals.ai ↗ |
| Claude Sonnet 4.6 Anthropic · senior seat | 92.1% vals.ai ↗ | 82.1% vals.ai ↗ |
| Claude Haiku 4.5 Anthropic · fast seat | 79.6% vals.ai ↗ | 81.2% vals.ai ↗ |
| Amazon Nova Pro Amazon · document analysis | 81.1% Stanford HELM ↗ | 73.6% Stanford HELM ↗ |
| Llama 4 Maverick Meta · open-weight diversity | 43.3% vals.ai ↗ anomalously low vs its own family on this harness | 77.8% vals.ai ↗ |
| Llama 3.3 70B Meta · independent reasoning seat | — closest published sibling (Llama 3.1 70B): 84.8% vals.ai ↗ | 77.2% vals.ai ↗ |
| Amazon Nova Lite Amazon · fast cross-check seat — cross-checks, never chairs | — no published score for this exact model | — no published score |
| Mistral Pixtral Large Mistral (EU) · European lab · vision seat | — its text backbone (Mistral Large 24.11): 76.2% vals.ai ↗ | — no published score |
Independent published evals of the exact model versions where available — siblings labeled as such, no score invented, re-verified 2026-07-03. Official benchmarks: MedQA paper · MedQA data · LegalBench · LegalBench paper. Independent leaderboards: vals.ai MedQA · vals.ai LegalBench · Stanford HELM · NCBE MBE. These are US-common-law and US-medical benchmarks — independent evals, which is exactly why this service stays triage-only. It tells you whether to hire a professional; it never replaces one.
What would this cost with a lawyer instead?
A records review by an attorney typically runs $250–$500/hour, and a careful read of a medical chart takes roughly 2–4 minutes per page. That means: 100 pages ≈ $1,500–$3,500 · 300 pages ≈ $4,000–$10,000 · 1,000 pages ≈ $12,000–$25,000 — before anyone tells you whether the case is even worth pursuing. Most firms won’t start without a retainer. A human-led expert panel (attorney + reviewing physician + medical-record analysts) runs $5,000–$15,000+ before an opinion lands. The Full Bench delivers its AI Opinion & Verdict on the same 1,000 pages for $249 — so you only spend lawyer money when the records say it’s worth it. Our system doesn't replace any of them — but it sure helps to educate you before you pay them. (Ballpark market rates; your region varies. This is triage, not legal advice.)
Flat fees. Larger or complex record sets may be quoted on intake. CaseCheck is an informational, 100% AI-generated case-strength assessment built on a family doctor's framework — it is not reviewed by a physician per case, is not legal or medical advice, and does not replace a lawyer or your own medical care.
Fill this in, upload your records, and the AI gets to work. You'll have your answer in days.
Generic AI gives a shapeless answer. CaseCheck runs a family doctor's medical framework as 100% AI — so you get a fast, structured, plain-language read on what your records actually support, affordably.
Designed by a family doctor; every assessment is generated by AI — fast and consistent.
Your records are handled securely and only used to answer your question.
Records in either language, patients anywhere — a clear answer wherever you are.
No to both. CaseCheck gives you an informational, 100% AI-generated assessment — built on a family doctor's medical framework, but not reviewed by a physician per case — of what the medicine in your records supports. It helps you decide whether, and to whom, to take your case. Only a licensed lawyer can give legal advice.
Not at all. You just upload them and tell us what happened in your own words. The AI reads them for you and explains it plainly.
Most case checks come back within a few days. Larger record sets may take a little longer; we'll tell you upfront.
Yes. Your records are handled securely and used only to prepare your assessment. We don't share them.
Injury and car accidents, medical malpractice, disability and work leave, denied insurance claims, and workers' comp. If you're not sure, just submit and we'll tell you whether it's something we can assess.
Then you've saved yourself the cost and stress of chasing a claim that won't go anywhere — that answer is worth having too. We tell you straight either way.
A few minutes now could save you thousands — or tell you the truth before you waste them.