How Accurate is PubGo?

Real performance metrics from our models, tested on articles we've never seen before.

71%

Top-3 Accuracy

Right journal in first 3 picks

#1.6

Avg. Position

Where the right journal ranks

6

Active Specialties

With trained models

0%

Hallucination Risk

No AI text generation

Performance by Specialty

Each specialty has its own trained model. Metrics are from testing on unseen articles.

SpecialtyStatusTop-3Tries Needed
Cardiology✓ Active62.70%1.78
Obstetrics, Gynecology & Reproductive Medicine✓ Active76.3%1.43
Pediatrics✓ Active81.4%1.35
Radiology, Nuclear Medicine and Imaging✓ Active70.5%1.64
Surgery✓ Active64.8%1.71
Health Informatics✓ ActiveTBDTBD
General Medicine & Family MedicineComing soon
ENTComing soon
Internal MedicineComing soon
OrthoComing soon
GastroenterologyComing soon
NeuroComing soon
Immunology & Infectious DiseaseComing soon
OphthalmologyComing soon
AgingComing soon
AnatomyComing soon
Anesthesiology and Pain MedicineComing soon
Applied Microbiology and BiotechnologyComing soon
Biological PsychiatryComing soon
Biomedical EngineeringComing soon
Clinical BiochemistryComing soon
Clinical DentistryComing soon
Clinical PsychologyComing soon
Critical Care and Intensive Care MedicineComing soon
DermatologyComing soon
Emergency MedicineComing soon
EndocrinologyComing soon
EpidemiologyComing soon
GeriatricsComing soon
Health PolicyComing soon
HematologyComing soon
HepatologyComing soon
Medical EducationComing soon
NephrologyComing soon
Nursing (Clinical)Coming soon
OncologyComing soon
Pathology and Forensic MedicineComing soon
Pharmacology & ToxicologyComing soon
Physiology & Basic Medical SciencesComing soon
PodiatryComing soon
Psychiatry and Mental HealthComing soon
Pulmonary & RespiratoryComing soon
RadiationComing soon
Radiological and Ultrasound TechnologyComing soon
Rehabilitation & Physical PerformanceComing soon
Reviews and ReferencesComing soon
RheumatologyComing soon
TransplantationComing soon
UrologyComing soon
📊 What do these metrics mean?

Top-3 Accuracy

In plain terms: If we give you 3 journal suggestions, how often is the "right one" in that list?

e.g., 94% means 94 out of 100 times, the journal where the article was actually published appears in our top 3.

Tries Needed (Submissions for Acceptance)

In plain terms: On average, how many journals do you need to try before finding the right fit?

A value of 1.5 means most people find their journal on the 1st or 2nd try using our rankings.

Aimed High / Aimed Low (Tier Shifts)

Aimed High (Oversell): We suggested a more prestigious journal than where the article ended up.

Aimed Low (Undersell): We suggested a less competitive journal than needed.

MRR (Mean Reciprocal Rank)

In plain terms: A technical score that measures how high the correct journal ranks in our list.

1.0 = perfect (always #1), 0.5 = usually around #2. Higher is better.

🔬 How does PubGo work?

PubGo analyzes your paper's title and abstract and compares them to millions of real, published biomedical articles. It identifies which specialties your work belongs to and ranks journals where similar papers are actually being published today.

This is NOT generative AI

PubGo does not make up text, rewrite your paper, or fabricate content. It's a statistical matcher trained on real publication patterns—similar to how search engines evaluate relevance.

  • Understands clinical vs. sub-specialty context, not just keywords
  • Recognizes patterns across thousands of journals
  • Routes articles to the specialties they truly belong to
  • Compares your work to what journals actually accept, not just their "aims & scope"
✅ Using PubGo responsibly

Use it as:

  • ✓ A starting point for journal selection
  • ✓ A reality check for unrealistic targets
  • ✓ A fast navigator of publishing options
  • ✓ A way to discover relevant journals

Don't use it as:

  • ✗ A guarantee of acceptance
  • ✗ A prediction of reviewer decisions
  • ✗ A replacement for reading journal scope
  • ✗ A reason to skip your own evaluation

Your expertise + PubGo = the strongest strategy.

Treat the Top Match as your "most statistically plausible home," then apply your own judgment.