Free AI Blood Test Explainer

Turn a small set of supported lab values into plain-language questions to discuss with a care professional. This tool explains terms; it cannot diagnose a condition, check a complete report, or decide what treatment you need.

Private-by-design input flow

Explain selected lab values

Paste a few supported rows. Nothing is sent until you submit after checking both consent boxes.

Free service limits: up to five requests per network address per 15 minutes, with a shared daily allowance. Attempts may count even if the AI service fails. Do not wait for this tool if you need medical care.

Lab values as text or CSV

The pre-filled rows are examples, not your results or universal reference ranges. Replace them with the values and units from your own report.

One row per test: test,value,unit or test,value,unit,low,high. Optional header: test,value,unit,low,high. Use plain delimited rows only; no names, dates or report identifiers.

Supported test IDs and exact units: hba1c (% / mmol/mol) · glucose (mg/dL / mmol/L) · hemoglobin (g/dL / g/L) · wbc (10^9/L / cells/µL) · platelets (10^9/L / k/µL) · ldl (mg/dL / mmol/L) · hdl (mg/dL / mmol/L) · cholesterol (mg/dL / mmol/L) · triglycerides (mg/dL / mmol/L) · tsh (mIU/L / µIU/mL) · creatinine (mg/dL / µmol/L). Enter 1–10 distinct tests; ranges must be provided together and in the same unit.

Select TXT or CSVMaximum 12 KB. Processed locally; no PDF, image or OCR support.

Parsed preview

Only these rows can be sent

Test	Value	Unit	Provided reference range
HbA1c	5.4	%	No range supplied
Hemoglobin	14	g/dL	12–16 · Within supplied range

I am 18 or older and am using adult lab values.I consent to sending only the parsed test names, values, units and any ranges shown above to the configured AI service for an explanation.

Important: not for urgent results

AI may be incomplete or wrong and cannot identify every emergency. For a critical flag or urgent symptoms, contact your care team or local emergency service now; do not wait for this tool.

Only parsed marker names, values, units and optional ranges are sent to the configured AI service (OpenAI through Replit). Names and report IDs are not sent, and the app does not save submitted values in its database. The provider’s processing policies apply; this is not a claim of zero provider retention.

The math behind it

Formula and assumptions

No clinical calculation is performed. The tool sends only validated marker names, values, units and optional laboratory reference limits for an AI-generated explanation.

A careful way to read a lab result

A laboratory result is one piece of a larger health picture. The reason a test was ordered, symptoms, medicines, age, pregnancy status, prior results and the laboratory method can all matter. A value marked high or low is not, by itself, a diagnosis; a result inside a printed interval does not rule out every problem. Reference intervals are statistical guides chosen by the laboratory, and they can differ between facilities and methods. Use the interval printed on your own report when you have it, rather than comparing a number with an unrelated website.

This explainer accepts a deliberately limited set of common markers and a plain text or comma-separated list. It does not read a PDF, photograph, scan, or arbitrary report narrative. Before sending anything, remove names, dates of birth, addresses, record numbers and other identifying information. The browser parses the rows locally and previews exactly what will be sent. You must confirm you are at least 18 and consent before a request is made. Only the parsed marker name, number, unit and any range you supplied are sent to the configured AI provider. The app does not save the submitted values in its database, but provider processing policies still apply.

Use the explanation as a conversation aid

The response may simplify nuance, misunderstand a unit, omit context or be wrong. Check that each displayed test and unit match the original report, then ask the clinician who ordered the test what it means for you. Useful questions include what prompted the test, whether a repeat or follow-up is needed, and whether medicines or recent illness could affect the result. Do not change a medicine, start a supplement, or delay care because of generated text.

The service is not an emergency monitor and cannot reliably identify every urgent pattern. If a report is flagged critical, or symptoms feel severe or rapidly worsening, contact the care team or local emergency service now rather than waiting for an online explanation. If a test is unsupported, the unit is unfamiliar, or the report uses a special method, leave it out and ask the laboratory or clinician. Keep the original report available: its notes and full context are not included in this small text-only tool.

Background reading

Sources and context

These references provide background for the formulas and limitations. A link is not an endorsement or evidence that the source reviewed this tool.

MedlinePlus: Laboratory tests

More context

Before you use this number

Does this diagnose a condition?

No. It gives general explanations of submitted terms and possible discussion questions, not a diagnosis or treatment recommendation.

What information is sent?

Only supported parsed marker names, values, units and optional reference limits. Do not include personal identifiers. The AI provider processes the request under its own policies; the app does not save submitted values in its database.

Can I upload a PDF or photo?

No. Paste values or select a TXT/CSV file up to 12,000 characters. There is no PDF, image reading or OCR.

What if my report says critical?

Contact your care team or local emergency service promptly. This AI explainer cannot detect every emergency and should never delay urgent care.