BastionGPT vs DeepScribe: AI Medical Scribe Comparison for 2026

Published pricing you can act on today

DeepScribe does not publish pricing. Its Terms of Use state that fees are set in an Order Form negotiated with each customer, and buying requires a sales consultation. Third-party pricing guides estimate $350 to $500 per provider per month, with some estimates ranging from $200 to $750 depending on volume, integration depth, and features, plus reported setup fees of $500 to $1,000 and annual contracts.

BastionGPT Professional starts at $20 per user per month, listed publicly. Professional Plus costs $45 and includes the full AI assistant, multi-document analysis, image and spreadsheet reading, and manual model selection. Prices have held steady for over two years, and annual billing includes a free month.

The math for a five-clinician practice, using the low end of the third-party estimates for DeepScribe: $350 x 5 x 12 comes to $21,000 per year. BastionGPT Professional Plus runs $2,700 ($45 x 5 x 12). That is a difference of $18,300 per year, and it grows to $27,300 at the high end of the estimate range. A solo clinician pays $540 per year for BastionGPT Professional Plus versus an estimated $4,200 or more for DeepScribe.

‍

A trial you can start today

DeepScribe is enterprise sales only. There is no self-serve trial: evaluation starts with a demo request, and deployment runs through contracting and implementation. That process fits a health system procurement cycle and rules out most solo and small-group buyers.

BastionGPT offers a 7-day free trial with full access to the AI scribe and AI assistant, no sales call required, and every paid plan carries a 45-day money-back guarantee. You can record your first session the same day you sign up.

‍

A full AI assistant, not just the visit note

DeepScribe's platform centers on the encounter: ambient scribing, AI pre-charting that pulls prior notes, labs, and medications from the EHR, and automated coding. It does not include a general clinical AI chat, and it does not analyze documents you upload.

BastionGPT includes an unlimited general AI assistant on every plan. Clinicians use it to draft referral letters, insurance appeal letters, treatment plans, and patient education materials; review documentation for incorrect names, pronouns, vague language, and suspected under- or over-coding; and analyze lab summaries and spreadsheets. BastionGPT achieved 99.3% sentence-level accuracy on clinical documentation tasks from a peer-reviewed study by Baylor College of Medicine and Texas Children's Hospital researchers, and reports that AI scribe users reclaim an average of 90 minutes per day.

‍

Document and spreadsheet analysis

On Professional Plus and Ultra plans, BastionGPT reads 10+ documents at once, up to 1,000+ pages of combined context. Supported formats include PDF, Word, Excel, CSV, PowerPoint, HTML, plain text, and JPG or PNG images, including scanned PDFs, charts, and graphs. A clinician can upload a full patient history, a psychological evaluation, and records from multiple providers, then query, summarize, and cross-reference all of it in one conversation. The Professional plan covers a single PDF, Word, or text document up to 30 pages.

DeepScribe approaches records differently: its AI pre-charting aggregates context from data already inside the EHR. That helps at the point of care, and it also means there is no path for uploading and analyzing outside records, insurance documents, legal correspondence, or spreadsheets. Excel and CSV analysis is not part of the product.

‍

A scribe that knows who is speaking

BastionGPT recognizes up to 10 speakers per session, live or from uploaded audio, with sessions up to 240 minutes. Speaker diarization separates the transcript by voice, and BastionGPT's AI automatically labels each speaker's name and role, for example clinician, patient, or parent. That matters for family sessions, group therapy, care conferences, and interpreter-assisted visits.

DeepScribe identifies the patient, clinician, and third parties in the room, but publishes no maximum speaker count and no session length limit, so buyers cannot verify capacity for group settings before signing a contract.

‍

AI models you can see and choose

BastionGPT names its engines: licensed and HIPAA-compliant versions of GPT-5.5, Claude, and Gemini 3 Pro. Every plan selects the best model for each task automatically, and Professional Plus adds manual model selection plus early access to new models as they ship. Naming the models lets you verify exactly what is reading your records.

DeepScribe builds on proprietary in-house models. Its published research describes HEAL, a 13-billion-parameter model based on LLaMA 2 and trained on clinical conversations, and the company reports strong internal benchmarks. It does not name the models that power production notes today, so note quality rests on a stack you cannot inspect or compare.

‍

Your data stays yours, and stays out of training

The two platforms take opposite approaches to training data. DeepScribe has publicly described processing and labeling more than 3 million de-identified clinical conversations captured on its platform to train proprietary models, including HealAI, and describes this dataset as the core advantage behind its AI. Customer encounters, once de-identified, feed model improvement.

BastionGPT never uses your inputs, documents, or audio to train AI models, and enters a signed Business Associate Agreement with every customer on every plan, including the free trial. BastionGPT supports HIPAA, 42 CFR Part 2, FERPA, PIPEDA, PHIPA, and Australian APP requirements, with PHI residency in the US, Canada, or Australia based on billing geography. Data is encrypted with AES-256 at rest and TLS in transit, runs on HITRUST CSF certified and SOC 2 Type II attested Microsoft Azure infrastructure, and is independently penetration tested at least annually. Retention stays in your control: chats persist until you delete them, documents are removed on deletion or after 30 days of account inactivity, and transcript retention is user-configurable from 1 to 365 days. In three years serving more than 10,000 healthcare organizations, BastionGPT has had zero breaches of protected health information.

DeepScribe executes a standard BAA with its customers as part of contracting, encrypts data with AES-256, and maintains a public trust center with a SOC 2 Type II report. Healthcare-specific frameworks beyond HIPAA, such as 42 CFR Part 2 for substance use disorder records or FERPA for school-based services, are not published on its site. For a practice handling SUD or school-based records, that difference is worth a direct question during procurement.

‍

Note flexibility on every plan

BastionGPT produces SOAP, Detailed SOAP, DAP, BIRP, H&P, and progress notes, and a Custom Note tab accepts any free-form prompt per recording: narrative formats, parent-friendly summaries, forensic documentation, and other specialty outputs. You can also upload example notes so BastionGPT matches your writing style. A Saved Prompts library with folder organization and team sharing is included on every plan.

DeepScribe generates notes from specialty-tuned templates, and its Customization Studio learns from clinician edits over time. Customization is a real strength of the product, and it is delivered through enterprise configuration rather than a per-recording free-form prompt you control from day one.

‍

Deployment paths from solo practice to enterprise cloud

BastionGPT deploys three ways. Individual clinicians and practices run it in the browser on any device with no IT project. Organizations that want to embed AI or build direct EHR connections use the public BastionGPT API, priced from $45 per month with deployment regions in the US, Canada, and Australia. Large organizations on the Ultra plan ($75 per user per month, starting at 100 users) can run BastionGPT within their existing enterprise cloud with SSO.

DeepScribe delivers as an AWS-hosted service and is available through the AWS Marketplace. A public developer API product or a customer-cloud deployment option comparable to BastionGPT Ultra is not published.

‍

User ratings on independent review sites

BastionGPT holds a 4.8 out of 5 average across the three major verified B2B review directories: Capterra, Software Advice, and GetApp. Reviewers most often cite time saved on documentation and the freedom to use AI on real cases because data protection is built in.

DeepScribe holds a 4.1 out of 5 on G2 across 26 verified reviews and a 4.3 out of 5 on Capterra across 8 reviews as of July 2026. Positive reviews praise reliability, responsive support, and reclaimed evenings and weekends. Critical reviews most often report transcriptions lost during server outages, findings placed in the wrong note sections, occasional speaker attribution errors, and notes stuck in processing when connectivity drops.

Ratings shift over time, so this page cites the source and date for each figure. The links in the sources section lead to the live numbers.

‍