Pinned Tweet
OpenMed 3.0 is out: Patient Journey.
One synthetic patient, five systems: a clinic note, hospital FHIR, an outside clinic's HL7, a lab CSV, a retina DICOM report. They disagree.
3.0 links every fact to its exact source, keeps the conflict open and amends without erasing.
Every systematic review starts with thousands of abstracts and one question. Which studies did the scan, the biopsy or the surgery?
New open model, OpenMed GLiNER2.5 ProceduresTests Small. F1 12.6 to 48.0 on 200 held-out abstracts. Apache-2.0.
huggingface.co/OpenMed/GLiNE…
Fine-tuned from @fastinoAI's GLiNER2.5 Small on 1,000 MedMentions abstracts, with three labels.
Lab procedures are its weak spot, F1 28.9, about one in four found. It also missed "maintenance hemodialysis" in the card's last line. Not yet tested on hospital notes.
48
10 people sent their first pull request to OpenMed during 3.3, and that makes us so happy!!
118 pull requests in ten days, 25 of them from 14 community contributors. Every name is on the card.
Good first issues are open if you want yours on the next one.
1
8
499
Good first issues
github.com/maziyarpanahi/ope…
Every PR and its author is in the release notes
github.com/maziyarpanahi/ope…
90
We tried to skip the review ten different ways. OpenMed 3.3 refused all ten.
Export before the drug name was confirmed, reuse an old sign-off after a correction, write FHIR from an unreviewed draft. Each came back with a reason. Only the reviewed note became FHIR R4.
1
14
745
Same synthetic dictation as Saturday's offline speech post. Changing "25 mg" to "25 milligrams" pulled two plan statements back to unreviewed until someone looked again.
Our script played the reviewer, and drafts stay under human review.
openmed.life/docs/multimodal…
122
The discharge summary looked clean. The patient's name was still inside, in white text, hidden text and a comment.
We hid a synthetic name in five files like that. OpenMed 3.3 found it in all five. Two came out redacted, two exports were stopped and one PDF was rejected.
1
10
677
The files and the patient are synthetic, built for this test, and the local PII screen is our open OpenMed PII SuperClinical Small (44M).
A comment is the case it won't guess on, so the export stops instead of shipping a half-cleaned file.
openmed.life/docs/multimodal…
157
We dictated a synthetic patient's name, birth date and phone number to OpenMed 3.3, fully offline.
All three came back masked before a single line of transcript was released. Three words it wasn't sure about, metoprolol, mg and a 2, wait for a person to confirm.
1
5
500
Whisper large-v3-turbo transcribes locally and our open OpenMed PII SuperClinical Small (44M) screens each line before release. Network blocked the whole run, 0 attempts.
One synthetic voice, one test. It shows the release gate, not recall.
openmed.life/docs/multimodal…
170
A clinic visit is a conversation, a scanned form, an ECG and an image. OpenMed 3.3 handles all four, locally.
Visit audio becomes a transcript screened for identifiers, then a FHIR draft a clinician reviews. Hidden text in a PDF is caught before export.
Out now, Apache-2.0.
pip install "openmed==3.3.0"
npm install [email protected]
Swift, Android and the service image ship at 3.3.0 too. Everything was tested on synthetic data, and every draft stays under human review.
github.com/maziyarpanahi/ope…
3
256
Asked to find anatomy in a sentence about cytokines, a general model returned tocilizumab. A drug.
Our new Anatomy model returned nothing. On 200 held-out abstracts its F1 went from 25.71 to 61.85.
Two new small open models, for drugs and anatomy:
"Arthroscopy of the knee showed a tear of the medial meniscus and loss of articular cartilage on the femoral condyle."
New open model: OpenMed GLiNER2.5 Anatomy Small. It finds all four structures. F1 25.7 to 61.9 on 200 held-out abstracts. Apache-2.0.
huggingface.co/OpenMed/GLiNE…
Fine-tuned from @fastinoAI's GLiNER2.5 Small on 1,000 MedMentions abstracts. It now finds 655 of 987 labeled mentions, up from 333, with 476 wrong tags instead of 1,270.
One label covers organs, tissues, cells and body fluids. Not yet tested on hospital notes.
love this!! they said make inkling your own, so shall we post-train inkling 2 small into the best open medical model a hospital can run in-house?
594B total, 17B active, reads images, 1M context, apache 2.0. even their own model card says medicine needs more fine-tuning first
Inkling 2 Small’s weights are now available for download on HuggingFace under an Apache 2.0 license.
Make Inkling your own!
huggingface.co/thinkingmachi…
"12 weeks of MTX combined with HCQ." Two drugs, both abbreviated.
New open model: OpenMed GLiNER2.5 DrugChemical Small. It finds both. On 200 held-out PubMed abstracts, its wrong tags fell from 1,399 to 678. 74M parameters, CPU, Apache-2.0.
huggingface.co/OpenMed/GLiNE…
Fine-tuned from @fastinoAI's GLiNER2.5 Small on 800 MedMentions abstracts, then scored once on 200 others: F1 45.1 to 67.4.
It finds the drug, not the dose, and it counts enzymes and receptors as chemicals. The card has 10 real outputs and its limits.
Same 1,723 Spanish disease mentions, two open models. mmBERT finds 18 more. GLiNER2.5 raises 193 fewer false alarms.
How we built both, the sealed test we opened once, where each one goes wrong and which one to pick:













