U.S. Legal Support alternatives, compared on what each vendor publishes
U.S. Legal Support publishes six of the eight facts a buyer needs and lists no price. The nine alternatives below are compared on the same eight rows: processing model, human QA, turnaround, pricing model, indicative cost, export formats, fit and rating. Of the eleven tools in the table, four publish a rate you can act on.
What U.S. Legal Support publishes
U.S. Legal Support is the incumbent shape of this market: a nationwide bureau where record retrieval, physician and nurse review, billing summaries and AI-powered summaries are ordered through one vendor. Its own FAQ explains why there is no price list: each request is unique. Retrieval runs 30 to 40 days depending on hold periods, and the Google review base averages 4.0 out of 5 across 123 reviews and fourteen offices.
Read the full U.S. Legal Support profile
Record retrieval (medical and other), medical record review/summaries prepared by physicians or nurses, medical billing summaries, chronological organ
U.S. Legal Support in the Content Hub →
Category: US Med-Legal Review / Legal Nurse Consulting · Medical Record Retrieval / ROI. Vendor claims captured 13 August 2026. Rows marked not published were searched for and not found on the vendor site.
11 tools, four questions each
Processing model, turnaround, pricing model, published rate. Follow any vendor name for the full eight-row profile. A gold tag means the vendor does not publish that fact, which is itself worth knowing before a call.
| Vendor | Processing model | Typical turnaround | Pricing model | Indicative cost | Published |
|---|---|---|---|---|---|
| U.S. Legal Support | AI-Powered Summaries + Human Record Retrieval | 30-40 days (record retrieval, hold-period dependent) | Not published | Not published | 6/8 |
| Compex Legal Services | Human-managed nationwide retrieval network | Published claim. Records retrieval: "about 15–20 days" via authorization, "around 30 days" via subpoena, on… | Flat fee / alternative pricing with prepayment options | Not published | 6/8 |
| American Retrieval Company | Human-managed retrieval + copy-service integrations | Average turnaround approximately 15 days | Flat/upfront fee with no-fee guarantee on failures | Not published | 6/8 |
| Record Retrieval Solutions | Human-managed retrieval + client tracking portal | Average 16 days (vs. HIPAA's 30-day requirement) | Flat fee per request | $45 flat fee per request (3-3,000 pages) | 7/8 |
| ChartSquad | AI + Human Team (AI summaries; human-team chronologies) | Published claim, repeated across the site: "Results in 15 Days" and "ChartSquad gets your results in about 15… | Per-page (AI summaries); per-request plus fee (retrieval) | $0.35/page AI summaries (~$350 per 1,000 pages) | 7/8 |
| Ontellus | Human-managed retrieval + proprietary technology platform | Most requests completed in under 30 business days | Not published | Not published | 5/8 |
| Magna Legal Services, LLC | Human-only / Manual Review (LNC-certified nurse reviewers) | Not published | Quote-based (tiered service options) | Not published | 5/8 |
| Steno | Human court reporters + AI transcript tools | Same-day delivery options available | Deferred payment (DelayPay, paid at settlement) | Not published | 6/8 |
| Wisedocs | AI + Human QA (expert clinician oversight) | Hours; case study cites turnaround cut from 14 days to 2 | Not published | Not published | 5/8 |
| Trivent Legal | AI + Human Expert Review (DepSum AI + physicians/nurses) | 5-7 business days | Per-page, tiered; billed only after work is completed | Summaries up to 200 pages start at $230 | 7/8 |
| Medrecords AI | AI drafts, a human decides. Every line cites its source page. | Minutes to hours per file | Pay per page, no subscription (Self-Service); annual licence for AI Enablement and On-Prem | from 10¢ to as low as 5¢ a page, duplicates free | 8/8 |
Of the 11 tools above, 4 publish an indicative cost. Across all 326 vendors profiled in the Content Hub, 57 do. Every claim was captured on 13 August 2026 from the vendor's own site; nothing here is a Medrecords AI rating of another vendor.
Four questions that decide this, and one that does not
Feature lists in this market converge within a quarter. What separates these tools is commercial and procedural, and most of it is visible before you take a call.
- Who signsSome tools produce a draft your own reviewer signs off. Others include reviewers of their own. Neither is better; they are different liabilities and different price points.
- What the meter countsPer page, per file, per user, per credit or per year. A seat licence and a per-page rate diverge fast once one file gets large.
- Whether output is checkableA summary you cannot trace back to a page is a summary you have to re-read the file to trust. Ask what a citation points at.
- Where the records goWhose cloud, whose BAA, whose keys. This is the question procurement asks and the one vendor sites answer least often.
- Not: the modelWhich foundation model sits underneath changes quarterly and tells you nothing about whether the output holds up on a file like yours.
Where Medrecords AI fits, and where it does not
The row above is ours, held to the same eight questions. So is this.
What it does
Reads the file you already have: PDFs, scans, handwriting, the DICOM imaging rather than only the radiologist's report. Removes duplicates, flags wrong-patient pages, builds the chronology, and links every line back to the page it came from.
What it costs
Ten cents a page on Self-Service, duplicates free, no seat licence and no subscription. AI Enablement and On-Prem are annual licences. All three are listed on the plans page, in numbers, without a call.
What it will not do
It does not retrieve records from providers, sign an opinion, score a case, or decide a claim. It drafts and cites; a qualified human decides. Any vendor promising otherwise is describing a liability, not a feature.
U.S. Legal Support alternatives: common questions
Run one of your own files against the table
Send a real file. You get the chronology, the duplicate report and the citations back, and you can judge the output rather than the marketing.
Test a file