Is Abridge pulling ahead of Microsoft in AI scribes?

Is Abridge actually overtaking Microsoft in AI medical scribes?

Abridge is now ahead of Microsoft in customer satisfaction and new AI scribe buying decisions, although Microsoft still has a larger publicly reported customer base.

The competition has changed considerably since Microsoft acquired Nuance for $19.7 billion in 2022. Nuance already had a strong position in medical speech recognition, giving Microsoft established hospital relationships and a natural advantage when AI-powered clinical documentation began taking off. Abridge entered the same market as a specialist, building software that listens to patient consultations and automatically creates clinical notes.

The clearest evidence of a shift comes from KLAS Research's October 2026 Ambient AI report. Abridge leads in buying decisions and ranks among the industry's innovation leaders, while Microsoft's customer-reported performance has fallen sharply.

That finding builds on an earlier market-share comparison from Menlo Ventures. In October 2025, the firm estimated Microsoft/Nuance controlled 33% of the $600 million ambient scribe market, compared with 30% for Abridge. The gap was already small.

Microsoft still has considerable reach. It reported more than 100,000 clinicians using Dragon Copilot and more than 600 health systems adopting its ambient technology in early 2026. Abridge, meanwhile, has expanded into more than 300 health systems.

Hospitals are increasingly favoring Abridge when making new purchases. Whether that has already translated into a larger total business for Abridge is harder to establish.

If you want more recent data on this point, please see our latest digital health market report.

Has Abridge already passed Microsoft in AI scribe market share?

Abridge was only three percentage points behind Microsoft in the last widely reported ambient AI market-share comparison, making a takeover entirely plausible, although newer comparable figures have yet to confirm one.

Menlo Ventures' October 2025 healthcare AI report estimated that the market had reached $600 million. Microsoft/Nuance accounted for 33%, followed by Abridge at 30%, Ambience Healthcare at 13%, and Suki at 10%.

Those figures put the two largest vendors remarkably close. Microsoft's estimated share represented around $198 million of the market, against $180 million for Abridge. That $18 million difference is a calculated share of estimated market spending, rather than either company's audited revenue.

Together, the two companies controlled approximately 63% of spending. That made the market fairly concentrated despite the number of smaller vendors entering it.

There is another useful observation in Menlo Ventures' research: customers were increasingly willing to reconsider their providers. Among large health systems surveyed, organizations using ambient scribes were roughly as likely to switch vendors over the following three years as to remain with their current provider. Among outpatient providers, 67% indicated a likelihood of switching.

So even a substantial installed base offers less protection than it would in a traditional hospital software market. With only three percentage points separating the leaders in the earlier estimate, relatively modest changes in purchasing could reverse their positions.

AI scribe vendor	Market share	Implied share of $600M market
Microsoft / Nuance	33%	$198M
Abridge	30%	$180M
Ambience Healthcare	13%	$78M
Suki	10%	$60M
Nabla	4%	$24M
Freed	4%	$24M
Other providers	7%	$42M

The figures come from Menlo Ventures' 2025 estimate. Percentages total 101% because of rounding, and the implied dollar amounts are calculations rather than reported vendor revenues.

Why are hospitals choosing Abridge over Microsoft Dragon Copilot?

Abridge currently has the stronger record in competitive AI scribe purchasing decisions. Hospitals increasingly prefer its product even when Microsoft is an established technology supplier.

KLAS Research identified Abridge as the leader in buying decisions in its October 2026 report. The same research placed Abridge, Ambience Healthcare, and Suki among the leading vendors for innovation, while highlighting a sharp decline in Microsoft's performance.

The trend had appeared earlier. In February 2025, KLAS gave Abridge a performance score of 95.1, compared with 91.6 for Microsoft DAX Copilot. Customers praised Abridge's results and cost-effectiveness, while Microsoft received positive feedback for reliability and its established documentation capabilities. Some Microsoft customers, however, raised concerns about pricing.

Abridge's 2026 Best in KLAS award, with a score of 94.7 out of 100, added another year of strong customer feedback.

The sequence is telling. Abridge first established a satisfaction advantage and subsequently emerged as the leader in buying decisions. That suggests favorable customer experiences are helping it win business.

KLAS does not publicly provide enough information to calculate exactly how many Microsoft customers have switched to Abridge. Nor does leading buying decisions automatically mean winning the largest contracts by dollar value. Still, Abridge's advantage in purchasing preference is supported by more than a single survey.

If you want more recent data on this point, please see our latest digital health market report.

Is Microsoft Dragon Copilot actually losing customers?

Microsoft Dragon Copilot is losing competitive momentum, but there is no public evidence of a widespread customer exodus.

In February 2026, Microsoft said more than 600 health systems had adopted DAX Copilot over the preceding 18 months. The company also reported that more than 100,000 clinicians were using Dragon Copilot in their daily practice.

Those figures give Microsoft a substantial established presence. Its Nuance business also brings years of experience selling medical speech technology to hospitals, where implementation and procurement can take considerable time.

The warning comes from customer feedback. KLAS's latest research describes a sharp drop in Microsoft's performance ratings as buyers demand more from ambient AI products. That raises questions about future renewals and purchases, particularly as competitors improve.

However, falling satisfaction scores do not tell us how many customers have canceled their subscriptions. Existing contracts can run for years, and hospitals may continue using a product even when they prefer an alternative.

There is also a difference between losing an entire hospital and losing part of its AI documentation budget. A health system might continue using Dragon Copilot in one department while introducing Abridge elsewhere.

Until Microsoft publishes comparable renewal figures or independent researchers measure actual vendor migrations, claims of a large-scale collapse in its customer base are premature.

Is Abridge growing faster than Microsoft in hospitals?

Abridge has expanded its hospital footprint remarkably quickly, but Microsoft's public disclosures are too different to calculate a reliable head-to-head growth rate.

In February 2025, Abridge reported deployments in more than 100 health systems. By August 2026, it was working across more than 300. The reported footprint therefore more than tripled in approximately 18 months.

Several deployments show the scale of its enterprise business. Johns Hopkins Medicine announced plans to make Abridge available to approximately 6,700 clinicians, while Duke Health extended access to around 5,000. Mayo Clinic expanded its deployment and began exploring nursing documentation.

Abridge also moved into federal healthcare. By September 2026, it reported deployments involving thousands of clinicians at more than 75 Veterans Affairs medical centers, through a competitively awarded pilot and a broader contracting arrangement.

These are demanding customers. Major academic medical centers and the VA have complex electronic health records, multiple specialties, and strict data-governance requirements. Abridge has clearly moved beyond small pilots.

Microsoft's disclosure of more than 600 adopting health systems provides a larger absolute reference point, although the companies do not use identical definitions of adoption and deployment. Microsoft's earlier count also cannot be used to calculate its subsequent growth.

Abridge is expanding fast. Whether its number of paid installations is now growing faster than Microsoft's remains unproven.

Does Microsoft still have more doctors using its AI scribe?

Microsoft has the stronger published clinician-user figure, with more than 100,000 Dragon Copilot users reported in March 2026.

Abridge has disclosed significant activity but has not published an equivalent company-wide active clinician count. In August 2026, it said its platform was expected to support more than 100 million patient-clinician conversations during the year.

That works out to roughly 274,000 conversations per day on an annualized basis. It's a substantial volume, but we cannot convert consultations directly into unique users. One doctor might record ten appointments daily while another uses the software only occasionally.

Microsoft's figure also requires some care. The company describes clinicians using Dragon Copilot as part of their daily practice, but it does not provide a standardized public measure of monthly active users, paid seats, or ambient-recorded encounters.

Intermountain Health offers a more concrete example. Microsoft reported over 2,500 active Dragon Copilot users at the health system in March 2026, showing that the technology is being used in everyday clinical work.

Abridge's deployments at large hospitals make it a credible challenger in user reach. For now, however, we have more complete evidence of Microsoft's clinician count.

Is Abridge's AI medical scribe better than Microsoft's?

Abridge currently has stronger customer satisfaction ratings than Microsoft Dragon Copilot, although independent research has not proved that its AI produces more accurate notes in every clinical setting.

The 2025 and 2026 KLAS evaluations consistently favor Abridge. These rankings draw on feedback from healthcare organizations about their experiences with the products, including outcomes, vendor relationships, and implementation.

That's useful information for hospitals making purchasing decisions. A tool that works well in a demonstration can still disappoint when thousands of clinicians start using it across different specialties and workflows.

But customer satisfaction is not the same as a controlled comparison of clinical accuracy. For instance, a hospital may prefer Abridge because it requires fewer manual edits, has better technical support, or costs less, even if both systems produce clinically acceptable notes.

Abridge also emphasizes Linked Evidence, a feature designed to connect generated documentation with supporting source information so clinicians can check the output. That is useful when physicians need to verify whether an AI-generated statement reflects the original encounter.

Microsoft has continued improving its documentation tools as well. Its Dragon products have a long history in medical dictation, and Dragon Copilot now combines ambient documentation with voice-driven workflows.

Abridge has the better customer experience record. Whether it also produces more accurate clinical notes across specialties is a different question, and one the available research hasn't settled.

If you want more recent data on this point, please see our latest digital health market report.

Do Abridge and Microsoft AI scribes actually save doctors time?

Both Abridge and Microsoft have shown that AI scribes can reduce documentation work, although the size of the improvement varies considerably between studies.

A Mayo Clinic primary-care study involving 332 physicians found that documentation time fell from 5.11 to 4.16 minutes per note after Abridge adoption. That's a saving of 0.95 minutes, or 18.6%.

Hawaii Pacific Health reported similar benefits in a study covering 79 providers and more than 25,000 generated notes. High-frequency Abridge users experienced a 21% reduction in daily documentation time, equivalent to 13.6 minutes.

Microsoft has positive evidence too. An observational analysis involving 210 DAX users found that active documentation time declined from 5.54 to 4.09 minutes per appointment, a reduction of 26.2%.

However, a randomized UCLA trial published in NEJM AI produced more cautious results. Researchers assigned 238 physicians across 14 specialties to Microsoft DAX Copilot, Nabla, or usual care. Nabla reduced note-writing time by a statistically significant 9.5% compared with controls, while DAX's estimated 1.7% improvement was not statistically significant.

The trial also showed how uneven real-world use can be. DAX was used during only 33.5% of eligible visits, Nabla during 29.5%, and around 15% of physicians assigned a scribe never used their assigned tool.

Abridge was not part of that trial, so its results cannot be treated as a direct victory over Microsoft. Still, the research provides a valuable reality check on the entire category. Positive observational studies do not guarantee equally large improvements in controlled experiments.

And if doctors barely use the tool, even excellent AI won't save them much time. Clinician adoption and workflow design clearly deserve as much attention as the underlying models.

Can Abridge beat Microsoft without owning the hospital's software?

Abridge has already shown that it can win major healthcare customers without controlling their electronic health records or cloud infrastructure.

This is especially important because Microsoft benefits from its ownership of Nuance and its established relationships with large organizations. Its healthcare customers may already use Azure, Microsoft security products, and Dragon speech-recognition tools.

Abridge has taken a different approach, integrating directly with the systems doctors already use. Through Abridge Inside, clinicians can access documentation capabilities within Epic environments such as Haiku and Hyperspace.

That matters because Epic is central to clinical workflows at many large US health systems. Hospitals are reluctant to introduce extra screens and manual steps into processes that already consume significant clinician time.

Microsoft also integrates with Epic, including through deployments such as Intermountain Health. So simply supporting Epic isn't enough to explain Abridge's recent success.

What we find more interesting is the customer overlap. Organizations such as Mayo Clinic, Duke Health, Johns Hopkins Medicine, and UPMC have selected Abridge despite having the scale and technical resources to evaluate major enterprise suppliers.

Abridge's customer record shows that hospitals are willing to buy specialized clinical AI software independently of their broader technology stack. Whether that advantage survives as Epic develops more native AI capabilities is a separate question.

Is Abridge becoming more than an AI medical scribe?

Abridge is rapidly turning its AI scribe into a broader clinical assistant, and its early usage figures suggest doctors are interested in more than automatic note-taking.

In April 2026, Abridge introduced context-aware clinical decision support with Wolters Kluwer's UpToDate, bringing established medical reference information directly into clinical workflows.

The system can use patient context to answer clinical questions, summarize records before consultations, and help prepare referrals or handoffs. These capabilities take Abridge beyond documenting what happened during an appointment.

By August, more than 300 partner health systems had adopted the decision-support offering, according to Abridge. Monthly active users represented more than half of eligible Abridge clinicians, while the number of queries per clinician had tripled over the previous two months.

At UPMC, chief medical information officer Rob Bart said more than 60% of Abridge users had adopted the decision-support functionality, largely without a separate training campaign.

The company is also expanding into payment and administrative workflows. Its June 2026 platform announcements included work with Availity on payer processes and AHIMA on coding-related capabilities, alongside initiatives involving life sciences and clinical research.

This could become a bigger competitive advantage than note generation itself. A product used for documentation, clinical questions, referrals, and administrative work becomes harder for doctors to give up.

There is still a gap between growing feature usage and proven additional revenue. But Abridge's reported query activity shows that its broader clinical-assistant strategy is being used, rather than merely announced.

If you want more recent data on this point, please see our latest digital health market report.

Can Microsoft Dragon Copilot catch up with Abridge's new AI features?

Microsoft has the technology and resources to compete with Abridge's broader clinical assistant, but its latest product expansion has yet to reverse Abridge's lead in customer preference.

At HIMSS 2026, Microsoft introduced additional Dragon Copilot capabilities focused on role-based experiences, hands-free workflows, and clinical intelligence beyond documentation.

The company's approach draws on several existing strengths. Dragon brings experience in medical speech recognition, Azure provides cloud infrastructure, and Microsoft's enterprise products support security and identity management.

Microsoft can also appeal to hospitals that prefer fewer technology suppliers. Cooper University Health Care, for example, cited Microsoft's security, compliance, scalability, and history of reliable software when explaining its selection.

Abridge is building around a somewhat different starting point: the medical conversation and the patient's clinical context. Its partnership with UpToDate gives clinicians access to medical evidence while they are already working in the patient's record.

The two vendors are now trying to capture much of the same opportunity, including documentation, clinical questions, and task automation.

Microsoft's advantage is the breadth of its technology and existing customer relationships. Abridge has the stronger reported product experience and has shown that clinicians will use new features when they are made available.

Microsoft can certainly develop similar capabilities. The harder task will be persuading hospitals that those capabilities offer a better everyday experience than the specialist products they already prefer.

Is Abridge actually cheaper than Microsoft Dragon Copilot?

Abridge has historically looked cheaper than Microsoft's premium ambient documentation products, but current enterprise contracts are too private to establish a reliable price comparison.

In its February 2025 research, KLAS reported that customers valued Abridge's cost-effectiveness, while some Microsoft DAX Copilot customers raised pricing concerns.

Sacra's analysis provides an indicative comparison. It estimated Abridge pricing at approximately $2,500 per clinician annually, against about $600 per month for premium Nuance DAX offerings.

That translates to roughly $208 monthly for Abridge versus $600 for the cited Nuance product. Abridge would be approximately 65% cheaper on those historical figures.

However, these figures are not verified quotes for equivalent current products. Microsoft's packaging has evolved with Dragon Copilot, and large health systems negotiate pricing based on seats, deployment size, functionality, and contract terms.

The economics also depend on actual use. An expensive product that saves several minutes per appointment could be worthwhile for a clinician seeing many patients daily. A cheaper product may offer poorer value if doctors spend considerable time correcting its output.

Both vendors now face additional pressure from EHR-native offerings, which may be included in existing software agreements or sold at a lower incremental price.

For the moment, Abridge's stronger cost-value reputation is supported by customer feedback. Its precise current pricing advantage is unknown.

Is Abridge making enough money to challenge Microsoft?

Abridge has already built a business with an estimated nine-figure recurring revenue base, giving it the resources to challenge Microsoft in clinical AI.

Sacra estimated that Abridge reached approximately $100 million in annual recurring revenue in May 2025, up from $60 million at the end of 2024. The firm also estimated contracted ARR of $117 million in the first quarter of 2025.

The distinction is important: contracted ARR can include subscriptions awaiting full deployment, while live ARR measures revenue associated with operating subscriptions. These are also third-party estimates, since Abridge is privately held and does not publish audited financial statements.

Abridge's financing followed a similarly steep trajectory. It raised $250 million at a reported $2.75 billion valuation in February 2025, then another $300 million at $5.3 billion in June. Sacra also records a $316 million Series E extension in April 2026.

The valuation nearly doubled in four months during 2025, indicating strong investor confidence in the company's growth prospects.

Microsoft, meanwhile, does not disclose Dragon Copilot's revenue separately. Comparing Abridge's ARR with Microsoft's total corporate revenue would tell us very little about their actual competition in medical documentation.

The relevant question is whether Abridge can convert its expanding customer footprint into recurring revenue while keeping implementation and AI processing costs under control.

There is enough evidence to establish a substantial commercial business at Abridge. We cannot yet determine whether that business generates more AI scribe revenue or stronger margins than Microsoft's competing product.

Could Epic and cheaper AI scribes threaten Abridge more than Microsoft?

Epic and lower-cost AI scribes could become a bigger long-term problem for Abridge than Microsoft, because hospitals may eventually consider basic automatic documentation a standard feature.

KLAS's October 2026 research found that EHR-native tools are changing expectations around integration and cost. Healthcare organizations are increasingly judging ambient AI as part of a wider technology purchase rather than as a separate innovation project.

That shift could favor Epic, which already controls the core clinical workflow at many large hospitals. If customers can access sufficiently accurate documentation directly within their electronic health records, they may become less willing to pay for another standalone subscription.

Other specialist competitors are also moving quickly. Ambience Healthcare and Suki joined Abridge among KLAS's innovation leaders, while Nabla and Heidi have attracted attention for different combinations of features and pricing.

Menlo Ventures' research adds a useful warning about the market's limits. Surveyed large health systems reported ambient scribe penetration of approximately 35% among clinicians, but expected that figure to reach only around 40% over the following three years. Those forecasts suggest that some institutions anticipate a plateau after the earliest adopters have signed up.

This creates two possible pressures: fewer clinicians left to convert within existing hospital customers, and tougher competition for each new contract.

Abridge's response has been to add decision support, coding-related tools, and other clinical workflows. Microsoft is following a broader-assistant strategy of its own.

Both are trying to give hospitals reasons to keep paying after automatic notes become commonplace.

What would prove that Abridge has definitively beaten Microsoft?

Abridge would have clearly overtaken Microsoft in AI scribes if it secured a sustained lead in paid usage and category revenue, rather than just a stronger position in new purchasing decisions.

Right now, the public evidence points in different directions depending on what we measure. Abridge leads in recent buying decisions and customer satisfaction. Microsoft has disclosed a larger installed health-system footprint and a six-figure clinician user base. The latest broadly comparable market-share estimate still puts Microsoft slightly ahead.

For a more conclusive result, we would want comparable disclosure of paid clinician seats, active usage, annual recurring revenue, contract renewals, and organizations switching vendors.

Those measures would also reveal whether Abridge's growth comes from winning new hospitals, expanding among existing customers, or replacing Microsoft inside organizations that use both tools.

Question we need to answer	What we know now	What remains unclear
Who has more market share?	Microsoft led 33% to 30% in Menlo Ventures' 2025 estimate	A comparable updated market-share figure
Who wins new purchases?	Abridge leads in KLAS's latest buying-decision research	Exact contract values and customer counts
Who has happier customers?	Abridge leads in KLAS customer performance	Long-term satisfaction across identical customer groups
Who has more users?	Microsoft reported over 100,000 clinicians	Comparable paid and active Abridge users
Who reaches more hospitals?	Microsoft reported over 600; Abridge over 300	Identical definitions and reporting dates
Who is growing revenue faster?	Abridge has substantial estimated ARR	Comparable Dragon Copilot revenue
Who retains customers better?	Public evidence is limited	Renewal, churn, and migration rates
Who saves doctors more time?	Both have positive studies; results vary	Large direct randomized comparisons
Who leads beyond documentation?	Both are adding clinical assistant features	Sustained usage, outcomes, and monetization

The biggest missing piece is comparable commercial data. Without it, even a strong customer-satisfaction lead cannot settle the question of overall market leadership.

So, is Abridge really pulling ahead of Microsoft in AI scribes?

Yes. Abridge is pulling ahead of Microsoft in the competition for new AI scribe customers, and its stronger satisfaction ratings suggest that this advantage is grounded in real product experience. Microsoft still has a larger publicly reported installed base, so Abridge's overall commercial lead remains unproven.

The sequence of evidence is particularly convincing. Menlo Ventures found the two companies separated by only three percentage points in estimated market share. KLAS subsequently ranked Abridge first for customer satisfaction and then identified it as the leader in buying decisions. Meanwhile, Abridge more than tripled its reported health-system footprint over approximately 18 months.

Its move into clinical decision support strengthens the argument. More than half of eligible Abridge clinicians were using the new functionality monthly by August 2026, according to the company, and query activity was growing quickly. That gives Abridge an opportunity to become useful to doctors well beyond writing medical notes.

Microsoft still has advantages that Abridge cannot easily copy: longstanding Nuance relationships, established enterprise infrastructure, and broad hospital distribution. Its Dragon Copilot product also continues to improve, and there is no verified evidence of mass customer departures.

For now, Abridge is winning the competition for purchasing preference. It has earned that position through customer satisfaction, major hospital deployments, and rapid product development.

The next step is translating that lead into a larger share of paid clinicians and recurring revenue. If Abridge manages that while Microsoft struggles to recover its customer ratings, the market-share reversal that once looked speculative could become the clearest measure of who won the AI scribe race.

If you want more recent data on this point, please see our latest digital health market report.