Introducing Sully API

Discover

Introducing Sully API

Discover

Notes that read
like you wrote them.

Before your first visit, Sully reads your past notes in the EHR to learn your format and style. Every edit teaches it more, so you change less each week.

Book a demo

CLINICAL NOTE

Subjective

Follow-up for lab review. Interrupted sleep and daytime fatigue over several weeks.

Objective

Alert and engaged throughout the visit.

Assessment

Sleep concerns and daytime fatigue discussed.

Plan

Review lab results with the patient. Clinician to finalize documentation.

Trusted by care teams at

You see the patient

while Sully writes the note.

Recording

12:48

You

Patient

Recording

12:48

You

Patient

Listens.

Records the visit in person or on a televisit, and tells your voice apart from your patient’s.

Assessment

Sleep concerns and daytime fatigue discussed.

Plan

Review lab results with the patient.

ICD-10 G47.9

CPT 99214

Assessment

Sleep concerns and daytime fatigue discussed.

Plan

Review lab results with the patient.

ICD-10 G47.9

CPT 99214

Assessment

Sleep concerns and daytime fatigue discussed.

Plan

Review lab results with the patient.

ICD-10 G47.9

CPT 99214

Writes.

Turns the conversation into a note in your template and your style, with ICD-10, CPT medical codes included.

Ready for your signature

Visit note

4 sections · Dr. Brandt

Diagnoses

G47.9 · Sleep disorder, unspecified

Orders

Lab panel · Follow-up in 2 weeks

Ready for your signature

Visit note

4 sections · Dr. Brandt

Diagnoses

G47.9 · Sleep disorder, unspecified

Orders

Lab panel · Follow-up in 2 weeks

Files.

Sends the note, diagnoses and orders to your EHR. You review and sign.

The model behind the note.

Fastest of the 8 models above 90% accuracy

Sully M1

92.2%

9.4 s

Claude Sonnet 5.5

92.1%

11.6 s

Muse Wonder

91.7%

14.8 s

Claude Opus 5.5

94.5%

16.6 s

Kimi K3

93.8%

23.7 s

GPT-6.1 Sol

92.7%

25.0 s

GPT-6 Astra

93.4%

30.1 s

Claude Fable 5.1

96.3%

50.4 s

50 s

Seconds per task · lower is better

Instead of renting a general-purpose AI, Sully Scribe runs on Sully M1, a model we tuned and run ourselves for medical tasks. It delivers top-tier accuracy at a speed and cost the big general models can’t match.

Explore the evaluation

92% accurate on clinical tasks

Fastest of the models above 90% accuracy

Never trained on your data

Fastest of the 8 models above 90% accuracy

Sully M1

92.2%

9.4 s

Claude Sonnet 5.5

92.1%

11.6 s

Muse Wonder

91.7%

14.8 s

Claude Opus 5.5

94.5%

16.6 s

Kimi K3

93.8%

23.7 s

GPT-6.1 Sol

92.7%

25.0 s

GPT-6 Astra

93.4%

30.1 s

Claude Fable 5.1

96.3%

50.4 s

50 s

Seconds per task · lower is better

Explore the evaluation

92% accurate on clinical tasks

Fastest of the models above 90% accuracy

Never trained on your data

Fastest of the 8 models above 90% accuracy

Sully M1

92.2%

9.4 s

Claude Sonnet 5.5

92.1%

11.6 s

Muse Wonder

91.7%

14.8 s

Claude Opus 5.5

94.5%

16.6 s

Kimi K3

93.8%

23.7 s

GPT-6.1 Sol

92.7%

25.0 s

GPT-6 Astra

93.4%

30.1 s

Claude Fable 5.1

96.3%

50.4 s

50 s

Seconds per task · lower is better

Explore the evaluation

92% accurate on clinical tasks

Fastest of the models above 90% accuracy

Never trained on your data

Median seconds per task and accuracy on Sully’s benchmark of clinical tasks drawn from real production work. Only models scoring above 90% are shown.

Built for the pace of clinical care.

Less time documenting. A fast draft, ready for your clinicians to review.

Note generation time

sec

Less documentation
time per note.

Note acceptance rate

94.3

Note acceptance rate

94.3

Clinical accuracy

92.7

Reliability

99.85

Note generation time

sec

Clinical accuracy

92.7

Less documentation
time per note.

Note acceptance rate

94.3

Reliability

99.85

Works inside the EHR you already use.

Supported integrations

BEFORE THE VISIT

Sully reads

The last 6 months of visits

Diagnoses already on the chart

Your EHR note templates

The appointment type, to pick the right template

AFTER THE VISIT

Sully writes

The note, section by section

Diagnoses to the structured list

Lab, referral and medication orders

ICD-10 and CPT codes

Why Sully.

Book a demo

Typical AI scribe.

The model doing the writing

Sully M1, our own model tuned for clinical work. 92% on clinical tasks.

Often a general AI model rented from another company.

How accuracy is checked

Every note is checked against the visit and the signed note. Every error is sorted by type and cause.

Rarely shared.

Learning your style

Learns from every edit you make. No ratings, no forms.

Often needs you to rate notes or change templates yourself.

Your EHR

Works with all major EHRs. Reads past visits and writes to your custom fields.

Often copy and paste, or a basic link.

Your data

Never used for training. No setting changes that. BAA with every customer.

Check the fine print for training opt-outs.

The model doing the writing

Sully M1, our own model tuned for clinical work. 92% on clinical tasks.

How accuracy is checked

Every note is checked against the visit and the signed note. Every error is sorted by type and cause.

Learning your style

Learns from every edit you make. No ratings, no forms.

Your EHR

Works with all major EHRs. Reads past visits and writes to your custom fields.

Your data

Never used for training. No setting changes that. BAA with every customer.

Typical AI scribe.

The model doing the writing

Often a general AI model rented from another company.

How accuracy is checked

Rarely shared.

Learning your style

Often needs you to rate notes or change templates yourself.

Your EHR

Often copy and paste, or a basic link.

Your data

Check the fine print for training opt-outs.

Your data is private

We’re not just leading in compliance and security, we also make sure all your data is private to you. We don’t use or train on your data.

ISO 27001

Information security management

HIPAA

Health data privacy & safeguards

HITRUST

Healthcare security assurance

ADHICS

ALIGNMENT

Abu Dhabi health cybersecurity

SOC 2

TYPE II

Independent controls audit

PDPL

READINESS

Regional data protection

Explore our Trust Center

Notes in seconds. In your style. In your EHR.

Book a demo

Notes that readlike you wrote them.

Notes that readlike you wrote them.

Notes that readlike you wrote them.

Notes that readlike you wrote them.

You see the patient

You see the patient

You see the patient

Listens.

Assessment

Plan

Assessment

Plan

Assessment

Plan

Writes.

Files.

The model behind the note.

The model behind the note.

The model behind the note.

The model behind the note.

Built for the pace of clinical care.

Less documentationtime per note.

Less documentationtime per note.

Less documentationtime per note.

Works inside the EHR you already use.

Works inside the EHR you already use.

Works inside the EHR you already use.

Works inside the EHR you already use.

Sully reads

Sully writes

Why Sully.

Why Sully.

Why Sully.

Typical AI scribe.

Typical AI scribe.

Typical AI scribe.

Your data is private

Notes in seconds. In your style. In your EHR.

Notes in seconds. In your style. In your EHR.

Notes that read
like you wrote them.

Notes that read
like you wrote them.

Notes that read
like you wrote them.

Notes that read
like you wrote them.

Less documentation
time per note.

Less documentation
time per note.

Less documentation
time per note.