Skip to main content
Beta Test

Find Your Rare Diagnosis in Minutes.

SecondLook is built for people navigating undiagnosed illness — the ones who’ve been told “I don’t know” or “it’s probably nothing” one too many times. For over 40% of cases, we name the correct diagnosis as the #1 match, so you can bring it to your doctor and ask — “could this be it?”

Get Your Analysis — Free
~10 minutes. No login. No cost.Have questions? Read the FAQ →
DIAGNOSTIC ANALYSIS
Outcomes

What SecondLook does for you

See the diagnosis your doctors may have missed.
over 40%
patients get the correct diagnosis as our top suggestion.
Walk into your next appointment with a clear next step.
6 in 10
patients get the diagnosis, or the single test that would confirm it.
Turn years of searching into a clear plan.
8 in 10
patients would reach an answer within our top 5 recommended tests.

Validated against Phenopacket2Prompt — the same rare-disease research benchmark used in Nature Biotechnology[1] and European Journal of Human Genetics[2].

Who it’s for

If any of these sound like you, we hope SecondLook can help.

Undiagnosed adults

You’ve seen multiple specialists over months or years and still don’t have an answer that fits.

Parents navigating a child’s illness

Your child has symptoms no one seems to connect. You’re looking for a way to think through the possibilities.

A diagnosis that doesn’t feel right

You have a diagnosis, but symptoms it doesn’t explain. You want a second look at the picture.

Important Medical Notice

This analysis is AI-generated and is for educational purposes only. It does not replace professional medical advice, diagnosis, or treatment. Always consult with qualified healthcare providers for medical decisions, especially before acting on any AI-suggested diagnosis or test.

How it works

How SecondLook works

Here’s what happens after you tell us your story: nine steps that turn it into a ranked list of possibilities and the specific tests that could confirm each one.

Each evaluation takes 8–10 minutes on average.

Step 1
Tell us your medical story and upload relevant history and data
Step 2
SecondLook extracts clinical concepts
Step 3
Map symptoms and concepts to candidates in our 9k+ rare disease knowledge base
Step 4
Activate 5 most relevant AI specialist agents in parallel to debate and select most likely diagnoses from profile
Step 5
Synthesize and rank a top-10 differential diagnosis list
Step 6
Refine diagnoses with 3–5 targeted patient questions
Step 7
Finalize the top-10 differential and probabilities
Step 8
Recommend tests to rule diagnoses in or out
Step 9
Deliver the final report
How SecondLook compares

Measured against the leading AI models

On real published rare-disease cases — the same benchmark used in peer-reviewed research evaluating diagnostic AI.

Already tried ChatGPT or Claude?

We’re not a chatbot — we’re a diagnostic pipeline built for rare disease.

Instead of asking one AI for an answer, SecondLook runs your case through the 5 most relevant AI specialists in parallel, grounded in a knowledge base of 9,275 rare-disease profiles, and scores each candidate against formal diagnostic criteria. On the same rare-disease cases, we’re 30–35% more accurate at the #1 match than a single query to OpenAI o3 or Claude Opus 4.7.

SecondLook
42.0%
correct at #1
OpenAI o3
31.5%
correct at #1
Claude Opus 4.7
30.9%
correct at #1
Methodology
  1. #1-match accuracy (over 40%) measured on a random sample of Phenopacket2Prompt cases (n=29). Independently re-verified across a larger 96-case random sample where SecondLook achieves 34% under the strictest paper-faithful Mondo grading.
  2. Head-to-head comparison against OpenAI o3 and Claude Opus 4.7: identical rare-disease vignettes, LLM tier grader applied uniformly across all three systems (n=50).
  3. “One test to the answer” and “five tests to the answer” measured on 96 random Phenopacket2Prompt cases. A test counts as confirmatory when a positive result would definitively establish the diagnosis under expert clinical judgment.
  4. Phenopacket2Prompt is a public benchmark of 9,587 published rare-disease case reports with verified ground-truth diagnoses (doi:10.5281/zenodo.15065293), used in peer-reviewed research (Robinson et al., European Journal of Human Genetics, 2026).
Guides

Health resources & rare-disease guides

Guides on rare-disease diagnosis, navigating complex medical cases, and making the most of AI symptom checkers on your health journey.

You’ve carried this long enough.

Ten minutes of your time. A ranked list of what might be going on, and the specific tests that could confirm each one. Free while we’re in early access.

Get Your Analysis — Free
Sources
  1. Jacobsen JOB, Baudis M, Baynam GS, et al. (2022). The GA4GH Phenopacket schema defines a computable representation of clinical data. Nature Biotechnology 40(6):817–820. doi:10.1038/s41587-022-01357-4.
  2. Robinson PN, et al. (2026). Evaluation of large language models on rare-disease diagnosis using the Phenopacket2Prompt benchmark. European Journal of Human Genetics. Benchmark data: doi:10.5281/zenodo.15065293.

#1-match accuracy and head-to-head comparisons measured on random samples of Phenopacket2Prompt vignettes, scored by an LLM tier grader applied uniformly across SecondLook, OpenAI o3, and Claude Opus 4.7. Full methodology footnotes appear inside the “How SecondLook compares” section above.