Operator: Good afternoon, and welcome to the Beta Bionics Q3 2026 earnings conference call. At this time, all participants are in listen-only mode. After the speaker's presentation, there will be a question and answer session, and instructions will follow at that time. As a reminder, please be advised that today's conference is being recorded. Now I would like to hand the conference over to Blake Beber, Head of Investor Relations. Please go ahead, sir.
Blake Beber: Good afternoon, and thank you for joining Beta Bionics' Q3 2026 earnings call. With me on today's call are Chief Executive Officer, Sean Saint, and Chief Financial Officer, Stephen Feider. Both the replay of this call and the press release discussing our Q3 2026 results will be available on the investor relations section of our website. Information recorded on this call speaks only as of today, 27 October 2026. If you are listening to a replay of this call, any time-sensitive information may no longer be accurate. Also on our website is our supplemental Q3 2026 earnings presentation and updated corporate presentation. We encourage you to refer to those documents for a summary of key metrics and business updates. Before we begin, we would like to remind you that today's discussion will include forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995.
Blake Beber: These statements reflect management's expectations about future events, our product pipeline, development timelines, financial performance, and operating plans, and speak only as of today's date. Please refer to the cautionary statements in the press release we issued earlier today for a detailed explanation of the inherent limitations of such forward-looking statements. These documents contain and identify important factors that may cause actual results to differ materially from current expectations expressed or implied by our forward-looking statements. We undertake no obligation to update them to reflect subsequent events or circumstances except to the extent required by law. With that, I would now like to hand the call over to Sean.
Sean Saint: Thanks, Blake, and thanks, everyone, for joining. Q3 is usually a somewhat boring quarter in diabetes, with summer vacations and a lull of news after ADA. Today, we are going to do our best to spice it up a little bit with some great updates. Let us dive in. In September, we received FDA clearance for Mint, our transformative patch pump, which we expect to fully commercialize in the US in Q1 of 2027. I feel this will ultimately prove to be the largest milestone achieved by the company to this point, and what an incredible effort it has been from the team to get that product designed, developed, and cleared by the FDA in a matter of about three years, which we believe is the fastest-ever timeline from concept to clearance for an automated insulin delivery, or AID, patch pump.
Sean Saint: We believe that Mint will be a game changer for people with diabetes. Starting in Q1 of next year, the only fully adaptive and most advanced insulin dosing algorithm in the US can power the smallest AID patch in the US, a patch pump that never requires recharging, integrates with industry-leading continuous glucose monitors, and can be controlled by iOS and Android smartphones. We designed Mint around the user experience, and the feedback we have gathered across our development program keeps reinforcing that. One recent example is from a survey of 30 active users of a competitive patch pump. All 30 rated Mint insertion as less painful. Delivering life-changing solutions that simplify and alleviate the burden of managing diabetes is critical. We strongly believe that Mint will deliver on that promise. In September, we completed a public offering, raising approximately $161 million in net proceeds.
Sean Saint: This capital strengthens our balance sheet and provides flexibility to support the commercialization of Mint, invest in our operations, and pursue future growth opportunities. Moving now to a brief overview of our Q3 performance. We delivered a record quarter of $35.5 million in net sales, which grew approximately 30% year over year. Q3 revenue growth was driven by growth in new patient starts and our growing installed base of users in the pharmacy channel, who we continue to retain at a high level. The percentage of new patient starts that were reimbursed through the pharmacy channel was in the high 30s. Our gross margin was over 57%, expanding 179 basis points year over year. Later, I will give more detail on our new algorithm, 3D Intelligence, and other projects in our innovation pipeline.
Sean Saint: First, I will hand the call over to Stephen to provide some additional detail on our Q3 performance and our updated full-year 2026 guidance. Stephen?
Stephen Feider: Thanks, Sean. Our revenue performance was mainly driven by new patient starts and the recurring revenue generated from our growing pharmacy install base. New patient starts increased a mid-single-digit percentage compared to the prior quarter, consistent with our expectations. We continued to expand the insulin pump market, as approximately 69% of our new patient starts came from people utilizing multiple daily injections prior to starting on iLet. A high 30s percentage of our new patients access iLet through the pharmacy channel, consistent with our expectations. Moving on to gross margin. Q3 gross margin was 57.3%, down 178 basis points quarter-over-quarter and up 179 basis points year-over-year. Gross margin was below our expectations, primarily reflecting a lower contribution from higher-margin DME iLet revenue due to distributor stocking dynamics. DME iLet sales, a high-margin revenue source, declined quarter-over-quarter, even as DME new patient starts increased.
Stephen Feider: Distributors supplied more iLets to patients than they purchased from Beta Bionics, which means the DME distributors ended Q3 with lower iLet inventory than at the end of Q2. This reflects normal variability in distributor ordering rather than weaker patient demand. At the same time, increased distributor stocking of lower-margin DME supplies, meaning infusion sets and cartridges, largely offset the revenue impact from DME iLets, but created an unfavorable mix for gross margin. We expect revenue mix to normalize in Q4 and are maintaining our full-year gross margin guidance. Meanwhile, continued growth and strong retention of our pharmacy installed base support a growing stream of high-margin recurring supply revenue. Total operating expenses in the third quarter were $46.9 million, an increase of 46% year-over-year. The increase in sales and marketing expense compared to the prior year is driven by expansion of our field sales organization and marketing investments.
Stephen Feider: The increase in R&D expenses relative to the prior year is driven predominantly by the Mint and bi-hormonal projects. The increase in G&A expenses relative to the prior year is driven by continued efforts to scale the company in support of commercial growth and pipeline initiatives. As of 30 September 2026, we had approximately $374 million in cash equivalents, and short and long-term investments. The capital raise was primarily intended to fund CapEx for Mint manufacturing and strengthen our balance sheet. We will continue to spend with discipline and expect to achieve positive free cash flow well ahead of historical diabetes peers. I'd now like to discuss our updated full-year 2026 guidance. We are increasing the lower end of our revenue guidance and now expect to generate total revenue of $124 million to $126 million, up from our prior guidance of $121 million to $126 million.
Stephen Feider: The updated guidance reflects our expectation to generate $29 million to $31 million of revenue in the fourth quarter. Our previous guidance reflected a potential revenue deferral impact of up to $10 million in the fourth quarter, as new prospective iLet users could choose to await the launch of Mint. While we are seeing some evidence of this occurring in the fourth quarter, it is not happening at the rate that we initially expected when Mint received clearance, and we now expect a potential revenue deferral impact of up to $8.5 million in the fourth quarter. On the percentage of new patient starts reimbursed through the pharmacy channel, we are reiterating our expectation of 37% to 39% for the full year. On gross margin, we are reiterating our full-year outlook of 58.5% to 59.5%.
Stephen Feider: This outlook reflects our expectation that the unfavorable revenue mix in Q3 will reverse in Q4, along with continued strong retention of our pharmacy installed base and the high-margin recurring supply revenue it generates. We expect a modest increase in operating expenses in the fourth quarter compared to the prior quarter, driven primarily by higher R&D expenses supporting our fully closed-loop pivotal study, which Sean will describe in more detail momentarily. We expect fourth quarter capital expenditures in the low to mid single-digit millions as we expand Mint manufacturing ahead of the full US launch in Q1 2027. Sean, back to you.
Sean Saint: Thanks, Stephen. I discussed our Mint update in detail earlier, and now I want to cover some other important updates in our innovation pipeline. First, 3D Intelligence, the next generation insulin delivery algorithm we unveiled in September. It is designed to give clinicians a choice of three algorithm options to better match each patient's individual insulin needs, while maintaining the simplicity and automation of our current algorithm. Following recent discussions with the FDA, we withdrew our 510(k) submission at the agency's request. During the review, we responded to FDA questions and requests for additional information. At the end of the 90-day review period, additional questions remained, and the agency asked us to withdraw and resubmit. While this is disappointing, we remain confident in 3D Intelligence potential. We are working to address the FDA's feedback and intend to submit the revised 510(k) in mid-Q4 2026.
Sean Saint: For clarity, the remaining questions were just that, questions on approaches that we took. We do not believe there were any concerns with the algorithm itself, and the resubmitted 510(k) is on the exact same algorithm with no changes, but with additional validation activities completed. Our current expectation is to launch 3D Intelligence in the US as iLet 3D and Mint 3D around mid-2027, subject to FDA clearance. Depending on the timing and resubmission and FDA clearance, a launch as early as the fourth quarter of 2026 remains possible. Together, these products would expand our automated insulin delivery platform, offering patients and clinicians greater choice in both algorithm options and device form factors.
Sean Saint: Regarding our real-world pivotal trial on adults with type 2 diabetes, we have completed enrollment and expect that study will enable us to expand our algorithm's indication for use to include adults with type 2 diabetes in the US around mid-year 2027, subject to regulatory clearance by the FDA. In Q3, we continued to see off-label prescribing of the iLet for people with type 2 diabetes, and we estimate greater than 30% of our new patient starts in the quarter were type 2. With the pivotal study completing enrollment, we are one step closer to potentially offering that community the means to access our algorithm on label. Another exciting update on the algorithm front is that in the coming weeks, we are expected to initiate a pivotal trial in the US studying our next generation fully closed-loop algorithm in adults with type 1 and type 2 diabetes.
Sean Saint: Our next generation fully closed-loop algorithm in development completely removes the meal announcement feature for users and represents a potential next step forward for the entire automated insulin delivery industry, and certainly the potential to further alleviate the burden of managing diabetes in a significant way. In the effort to maximize choice for our users wherever we can, we continue to advance our integrations with leading continuous glucose monitors. In Q4, we expect to launch our integration of iLet with Senseonics Eversense 365 and are committed to integrating Mint with Eversense 365 in the future. We continue to refine our bi-hormonal system. The drug reformulation is complete, and additional phase IIa trials are in the works. Assuming these trials proceed as planned, we believe the program is back on track, and we will continue to provide updates as we progress.
Sean Saint: We communicated a lot of updates today, and the collection of those updates go to show the foundation we have built and the steps that we are continuing to take toward building a category-leading diabetes technology company. This company is executing across the board at the highest level, and we have the team and the roadmap to deliver for the diabetes community. With that, I would like to thank you all for joining today's call, and we will now open it up for Q&A.
Operator: Certainly. Ladies and gentlemen, if you do have a question at this time, please press star 11 on your telephone. If your question has been answered and you would like to remove yourself from the queue, simply press star 11 again. Our first question for today comes from the line of Matthew O'Brien from Piper Sandler. Your question please.
Matthew O'Brien: Afternoon, thanks for taking the questions. I guess, Sean, maybe for starters on 3D, that is probably going to get the most attention today. Can you elaborate a little bit more on some of the additional questions that they had? Was it simplistically just validation testing they wanted you to do, or because it is three different algos or three different options for the patient. Is it more just patient-specific kind of testing, making sure they know what they are doing? Just anything more you can talk about as far as the delay there, and then I do have a follow-up.
Sean Saint: Yeah. Great question, Matt. Thank you. Your last portion of question there alluded to the fact that it may be like a human factors type thing, and that is not it at all.
Matthew O'Brien: Yeah.
Sean Saint: Primarily for us in this particular submission, simulation is a huge part of it. Simulation is a relatively new area for diabetes algorithms. We are all getting a lot better at it. The FDA offers a guidance document on what they believe they want for a simulation. Like many guidance documents, we are all still learning to interpret it perfectly. It is around that. It is around the validation of our simulator, not the results of it, not the algorithm itself, not what we do, but just ensuring that we can trust the output of the simulator. Again, not proposing that there is a problem. Instead, more proposing that we want to make sure that there is not. Hopefully, that is helpful. Again, it is not a human factors thing, it is not an output thing, it is not an outcomes thing. It is none of those things.
Sean Saint: It is just ensuring we did the right work to validate our simulator.
Matthew O'Brien: Got it. Thanks for that. I really do not want to make a mountain out of a molehill here, but the commentary that Stephen had on deferrals coming in lighter than expected, you cannot overread that as, "Hey, you know what? There is not as much interest in the patch as maybe we thought, and that is why we are not seeing as much deferral." Like, "Hey, you guys are a durable pump company. We like you guys. We like your algo." Is there anything to read into that commentary? Thank you.
Sean Saint: No, that is a really good question, too. I think it reflects a little bit more on an understanding of exactly how informed people are. Our fear was based on, of course, we have done market research, and we have asked our users who would prefer Mint, et cetera. Frankly, we are, I will say, very surprised and at some level scared by the answers and that people love the Mint product. If everybody had perfect information, then likely we would have seen more waiting. But the reality is that most people are not wandering through their lives wondering when Mint is coming out, right? They are just not that familiar with it. Because the awareness of the product is a little lower, the pauses were also a bit lower. But again, I do not mean that in a negative way.
Sean Saint: These are people who have not seen their healthcare providers, and certainly their healthcare providers are not in the business of pushing Mint at this point. But it was very hard for us to know who would know about it, who was following along, et cetera. We intended to be conservative on that.
Matthew O'Brien: Got it. Thanks so much.
Sean Saint: Yep. Thanks, Matt.
Operator: Thank you. Our next question comes from the line of Lawrence Biegelsen from Wells Fargo. Your question, please.
Lawrence Biegelsen: Good afternoon. Thanks for taking the question. First, I just wanted to start on Mint. When do you expect the first commercial shipments of Mint to occur? Do you need to wait for 3D to be cleared? How should we think about the ramp through 2027? I have one follow-up.
Sean Saint: Larry, I think we have said probably all we are prepared to say on when the first commercial shipment will be, just in the first quarter here. No, I do not think we need to wait for 3D, although the sooner we get it, the sooner we will be able to launch 3D on Mint, which is obviously a big deal for us just because we have a huge degree of confidence in each one of those products. They are independent, so we can ship without it. Great question on the ramp throughout the year. It is not exactly like this, but I think a good mental model is that we start, let us just say, near zero, and we add capability throughout the year.
Sean Saint: From a modeling perspective, model it as basically linear from now through the end of the year, averaging out at what we previously stated, which is at least 1.5 million units for the year. Is that helpful?
Lawrence Biegelsen: That is very helpful. Just for my follow-up, when Mint was approved, you lowered the 2026 guidance by $10 million. That is obviously $8.5 million now, and you expect to recover that next year. I think consensus was about $180 million when Mint was cleared. So I just want to ask the obvious question. It sounds like you are comfortable with about $190 million in sales next year. Can you confirm that? Just talk about maybe the shape of revenues next year. Thank you.
Stephen Feider: Yeah. Reiterating what you just said, Larry, we have communicated that any deferral that we would see from prospective Mint users who would have otherwise purchased iLet in 2026, we expect to more than make up for that deferral in 2027. So, your math is accurate. So now we are expecting the benefit into 2027 from 2026 deferral for the at least $8.5 million from where we currently have the consensus.
Lawrence Biegelsen: Got it.
Stephen Feider: Or where we had the consensus before Mint approval, I should say.
Lawrence Biegelsen: Got it. All right. Thanks so much.
Operator: Thank you. Our next question comes from the line of Chris Pasquale from Barclays. Your question, please.
Chris Pasquale: Yeah, thanks for taking the questions. I was hoping you could talk a little bit about your fully closed loop algorithm and what we might see there that differs from what other companies in the space are working on. Are you guys approaching that project differently, do you think, given your history of having a more kind of hands-off algorithm to begin with? I would just love any comments there on how you are taking maybe a little bit of a different angle.
Sean Saint: Yeah, Chris. Okay, let me take a moment to define a term here, which we do rarely. Hybrid closed loop, of course, means including interaction from the patient. Fully closed loop means not needing to interact, to require anything from the patient, meaning meal announcements, for example. Then there is a concept of forced fully closed loop, which is probably not a marketing term that we will continue with, but it is descriptive. What it means is not being able to meal announce, right?
Sean Saint: Purely a fully closed loop system. That is the direction we are going with this particular product. It is based on sort of a subsegment of the 3D algorithm, and importantly, it is going to be studied in both type 1 and type 2 diabetes. Said another way, we do believe that there is a huge cross-section of people that can benefit from an algorithm like that.
Sean Saint: And part of that, by the way, is in the primary care space, where it is more difficult to train on some of these things. So training becomes quite a bit easier when you are not forced to discuss, at any level, meal announcements, when the product just handles it. So I think when you compare the different development paths that different companies have followed, they are either a subsegment of users or they are maintaining a lot of the complexity of the previous device. We have tried to take a slightly different perspective, which is to include everybody, type 1 and type 2, while making this massively simpler device that can be used by anybody in any position. Said another way, we think it is the most commercially viable of the different architectures that are being studied right now.
Sean Saint: Now with that being said, that is not for everybody, and I think what we have seen lately from Beta Bionics is in terms of 3D, in terms of forcefully closed loop, et cetera, we have a pretty robust algorithm development team here in terms of our type 2 submission that we are talking about. And this is not our last bite at the apple.
Chris Pasquale: That is helpful. And then I was just hoping you would comment on the type 2 trial and what we should expect from here. When could that data be available? Is that going to line up with ADA next year, or could it come sooner when we would actually see that readout?
Stephen Feider: Yeah. The first time we will plan to communicate the data readout from the trial, again, acknowledging this is a real-world observational trial, is at the time when we get the clearance, which will be simultaneous with us having the label and then in effect, commercially launching it. So I cannot be more specific on your timeline relative to ADA other than what the public guidance that we have communicated, which is we expect this in the first half of next year.
Chris Pasquale: Okay. Thanks.
Operator: Thank you. Our next question comes from the line of Mike Kratky from Leerink. Your question please.
[Analyst] (Leerink): Hey, everyone. This is Sam on for Mike. Maybe just turning back to the Mint adoption ramp, could you just provide some more color on the degree to which capacity is currently in place to produce the 1.5 million units you expect? Any meaningful additional commercial milestones that need to take place ahead of the launch to really hit the ground running, to what degree you could potentially expand beyond the capacity, should outsized demand arise?
Sean Saint: Yeah, all good questions.
Stephen Feider: Yeah, all good.
Sean Saint: Let me start with the current capacity. As I answered earlier, I think with Larry, for modeling purposes, we should assume sort of near zero at this stage. But the reality is not quite near zero, of course. Well, it is effectively zero because the line is not up, validated, and ready to go. It is extremely close to that. We will start with a baseline level of production with a single manufacturing line. I will give one additional piece of color here. We are hiring for that line now. We are, I am not going to give an exact number, but meaningfully all the way through hiring a full four shifts for that line. Four shifts are what we believe are required to run the line effectively 24/7. Again, to keep that line up and operating at that level, we are near there.
Sean Saint: Opportunities for expansion beyond that, which are required throughout the year, are to replicate that line additional times. I am not going to state how many or talk about cycle times today. Opportunities for outperformance are primarily replicating additional lines. There are all your little ankle biters that pull away from your production capacity, right? Those are yield. Those are uptime of the line. Production capacity, meaning cycle time, which I am not going to state, but if we can ever improve cycle time of the line, that obviously improves output of each and every line. So those are the kinds of things we will be looking at in order to improve production capacity should it be necessary.
Stephen Feider: Okay.
[Analyst] (Leerink): Awesome. I just had a quick follow-up maybe on that point. Can you just provide a little bit more color on Mint's expected gross margin impact in 2027? Obviously, I think you mentioned it was going to be dilutive, but just the volume milestones needed to achieve accretion and when you expect that could occur.
Stephen Feider: Perfect. Yes. Mint's gross margin will be dilutive in 2027. Where we have the analyst models currently, the sell-side analyst models currently, we are comfortable with, and I won't give any specific guidance on 2027 yet, but just acknowledging again that we've communicated our expectations with sell-side analysts, and we have the models in a reasonably good place for 2027. Over the next three years, and this is the exciting thing, over the next three years, the cost characteristics of Mint and the design decisions that we've made for Mint allow us to have within three years, a gross margin outcome for that product that we expect to exceed our current patch pump incumbent.
Stephen Feider: And again, at the scale that we're talking about at that particular time period, the reason that we have the ability to achieve that level of gross margin or that gross margin profile, is because of the design decisions, meaning we have a durable component, which is where the meat of the product lives, sort of the expensive components that last two years. And then the disposable component, which we call Mint, that patients use whenever they're done with their 200-unit cartridge or up to three days plus 12 hours. That is much more cost-effective to make than what we're familiar with in the patch pump world. So super exciting gross margin profile and profit outlook for Beta Bionics in that regard, and I appreciate you asking about it.
[Analyst] (Leerink): Awesome. Thanks, guys.
Operator: Thank you. Our next question comes from the line of Michael Polark from Wolfe Research. Your question, please.
Michael Polark: Hey, good afternoon. Couple follow-ups on 3D Intelligence update. Just on timing quickly so I understand it, you are submitting more data back to FDA. It could be still possible you get that approval by the end of this year. I just want to make sure I understand that. That is only, say, 45 days. Sounds quick. The bigger question was, just about the validation of the simulator. Is it specific to the more aggressive or the more conservative versions of the algorithm, or it is not about one of the three at all? It is about how you tested and framed the system. I wanted to just understand if it was about one nudge of the algorithm or not. Thank you.
Sean Saint: Yep. Great questions, Mike. So, yeah, that would be quick. There is no question about it. We did announce in the prepared remarks that we will be resubmitting this year. We were very far through this submission, and unfortunately, we just ran out of time. So there is a possibility that the agency sort of picks up the new submission and looks at it as a continuation and picks up right where they left off. If they were to do that, there is, I would say, even a decent possibility that we could be looking at a clearance this year. They are not by any means obligated to do that, and the statutory numbers absolutely put us into next year. So that, I think, is the most likely scenario and what we are preparing everybody for with our guidance.
Sean Saint: But, if they were then to turn around and give it to us this year, I did not want everybody to look at us and say, "Hey, what gives? You said it was Q4, then it was H1, now it is Q4 again." Technically, that remains on the table. So we wanted to be honest about that. But again, the guidance itself is that we are likely to see this next year. On the second half of your question, the questions were very much specific to the simulator itself. They have nothing to do with individual modes of the algorithm.
Sean Saint: So I want to be clear. It is not that the simulator said, "Here is the output," and they said, "Ooh, that is concerning." That is not it. For any one particular mode, that is not it. What it was, "Hey, really interesting simulator, guys. You haven't provided enough data for us to believe it yet.
Sean Saint: We like what you showed us. We like the outcome, but we do not believe it yet." That is what validation of the sim itself is about. So again, the outcome of the simulator, the outcome of the modes, 3D itself, all that is good, we believe. But that is all hinged upon our simulator being believable, which is what validation of the sim is about. So no, not specific to modes, not even specific to the simulator, specific to the way we validated the simulator.
Michael Polark: Thanks, Sean.
Sean Saint: Yep.
Operator: Thank you. Our next question comes from the line of Steve Lichtman from William Blair. Your question, please.
Steve Lichtman: Thank you. Hi, guys. Sean, can you talk about how you see pharmacy coverage building in 2027 for Mint? I think you've said you'll start at the high 30s, 40% pharmacy mix you're at today. What are your latest views on what happens with prospective customers who don't get pharmacy coverage as well? Are you still assuming most purchase iLet?
Sean Saint: I'm going to kick that one to Stephen, actually.
Stephen Feider: Yeah. Hey, Steve. Pharmacy coverage expectations for 2027 for Mint are at the time of launch, we expect roughly 40% of all lives in the country to be covered for Mint, growing to around 50% by the end of the year. That is just the crisp way of communicating what expectation we have.
Steve Lichtman: For those that do not have pharmacy coverage and therefore will not be able to get Mint, are you still assuming most will purchase iLet, or what are your latest thoughts on that?
Stephen Feider: Well, there are a few different permutations of this, all of which will likely happen. Certainly, people who would prefer Mint who do not have insurance coverage for Mint, in some cases are going to buy iLet. I would actually use a data point that we even have internally today to prove that our particular value proposition of iLet will continue to win with many patients who would prefer to wear a tubeless pump. We actually surveyed several hundred or, yeah, a little over 300 of our users and asked them a question of, "If knowing what you know about Mint and assuming no difference in your out-of-pocket cost each month, would you prefer to switch to Mint right now?" And almost 90% of those patients said yes.
Stephen Feider: Now, that is a relatively small sample, but I think it proves that what we are selling, meaning the iLet's differentiation of a fully automated system, will continue to win even up against other patch pump form factors. So I think that people will certainly fall into the category of, "Okay, I am not covered for Mint, and I suppose I would prefer to be on a Beta Bionics product, and if I could choose, I would prefer Mint, but I am still going to purchase iLet." There will certainly be some of those people. Will there be some deferrals of patients who will otherwise wait to get Mint before they start on iLet? Absolutely. Quantifying that right now is a little bit difficult, candidly.
Stephen Feider: We have had FDA clearance for our product for 5, 6 weeks, and we have good experience in understanding what the patient demand is like for iLet in spite of there being Mint clearance. I think that will be different, of course, when we actually launch it commercially, rather than it just existing as a future product. Sorry to not give you a precise mathematical answer on that. I have a hard time not doing that as a finance guy, but I do not exactly know how each of those categories will get bucketed.
Steve Lichtman: No, I appreciate the color, Stephen. Just quickly, any update on the fully automated manufacturing timeline?
Sean Saint: Probably no update on the timelines for that. I suppose I will say that as we have discussed in the past, that we have a line up and operating. We have additional lines being constructed at this time, and we are working at this stage on our more fully automated manufacturing lines for the future. I think we have probably said in the past that was a future item. We are thinking about that now and doing the actual work now. But I am not going to provide exact timelines on it.
Steve Lichtman: Got it. Thanks, Sean.
Operator: Thank you. Our next question comes from the line of Jeffrey Cohen from Ladenburg Thalmann. Your question, please.
Jeffrey Cohen: Oh, good afternoon. Thanks for taking the question. Just two brief ones. I missed it, Stephen. Q3 pharmacy composition was what percent that you called out earlier in the call?
Stephen Feider: A high 30s percentage of our new patient starts went through pharmacy in the third quarter.
Jeffrey Cohen: Got it. When it comes to the Mint line, the $46 million CapEx that you called out, was that specific to one line? The anticipation, it sounded like that would be somewhat linear as Mint rolls out and rolls up as opposed to a step function. Any further color on that?
Stephen Feider: Yeah. The low to mid-single-digit CapEx that we will have in Q4 and then much of the CapEx that we will spend in 2027 is associated with our automated or semi-automated manufacturing line. That is actually more than one line. When I say line, I mean it is our semi-automated and manual manufacturing operation. Yeah, it is not one individual line. There is a separate project that Sean just kind of alluded to a moment ago of making the manufacturing for Mint more fully automated, and that has its own CapEx profile that I can get into if you would like, but that is more in future years, although we will start to see some of that CapEx incurred in 2027. Yeah. I think that answers your question.
Jeffrey Cohen: Okay. Most of this commentary is related to Mint, not specific to iLet, when you are talking about CapEx and automation.
Stephen Feider: That is right. Yeah. The CapEx for iLet manufacturing is behind us.
Jeffrey Cohen: Got it. Perfect. Thanks for taking the questions.
Stephen Feider: Yeah.
Operator: Thank you. Our next question comes from the line of Jeff Johnson from R.W. Baird. Your question, please.
Jeff Johnson: Yeah. Thanks. Guys, good afternoon. Maybe two algo questions, if I could, Sean. One, just on type 2, if you are using real world data there and you will not have the 3D algos out until later this year or probably early next year, will you get type 2 on 3D or how are you going to restrict use to just the standard iLet algo in the type 2 adult patients that you are targeting for mid-year approval there? Just the mechanics of how that type 2 approval will come across from an algo perspective.
Sean Saint: Yeah, that is a really good question, Jeff. Frankly, we are not exactly sure yet. As you have alluded to, our real world evidence study obviously studies people, type 2 users using the current algorithm as it sits today. That is one of the modes available in 3D because we know and love it, and users know and love it. So at a minimum, we would expect that we can get type 2 indication for that mode of the algorithm. But remember that we are also getting 3D, well, we hope to be getting 3D cleared primarily through simulation data. Assuming that that is the case, we may be able to apply the same level of thought to get this in a wider, fully, on all of 3D, if you will. But to be 100% honest with you, we are not sure yet.
Sean Saint: This is going to be a conversation that we have to have with the agency once we get into it.
Stephen Feider: One thing I need to add here. I was answering a type 2 question earlier that Chris Pasquale asked, and I just fact-checked myself looking at my notes. I had communicated in the response to the question that the type 2 indication was a H1 timeline. We are calling that a mid-2027 indication and commercial launch for type 2, not a H1 timeline. Go ahead and ask the second part of your question.
Jeff Johnson: Yeah. The second question is just on the forced FCL decision. I am a little surprised you are doing that on the T1 side, I guess. I mean, Sean Saint, you know more than any of us the alpha response there issues and concerns that if you cannot interact with your pump at all, does that increase in T1 at least the risk of hypo in a fully closed loop or a forced fully closed loop? I guess my question is, what is going to be your bar on what is acceptable from an increase in hypo risk from baseline to forced FCL or just any other thoughts you can give us on doing the forced FCL pathway in T1?
Sean Saint: Yeah. The first is, I do not think that forced FCL should result in a higher level of hypos. As with anything, diabetes is a tough beast. There are always going to be individual users that have one experience or another. But I do not think, as a general statement, forced FCL should be a higher level of hypo. In fact, I think it should probably be lower. But additionally to that, there is unquestionably an enormous group of type 1s today that are not getting access to AID generally, that absolutely has lower risks of hypo than MDI, that absolutely has lower risks of hypos than what you might call standard of care.
Sean Saint: And we need to make sure that they have access to that. And by providing a simpler system that is both simpler to use, simpler to train, simpler to understand from the healthcare provider standpoint, we are doing exactly that. Paradoxically, the reason we are doing this is actually to limit hypo in the population, to improve outcomes in the population. And we do that by getting products like this out to more people.
Jeff Johnson: Understood. Thank you.
Sean Saint: Thank you.
Operator: Thank you. Our next question comes from the line of Jonathan Block from Stifel. Your question, please.
Jonathan Block: Great, guys. Thanks. Good afternoon. Maybe just a couple of quick cleanups, then one on 3D. In terms of the cleanups, Stephen, are we now thinking new customer starts down, call it slightly less than 30%, Q3 to Q4 sequentially? I might have missed it, but did you have any timing on FCL in terms of approval? Then I will just ask the one on 3D.
Stephen Feider: Yeah, John, more like down 20% new patient starts in Q4 relative to Q3.
Sean Saint: This is approval timeline FCL. We haven't provided timing on that as of yet. Just start timing for the trial, which is relatively imminent. I don't remember the exact wording that we used.
Jonathan Block: Okay. Then just on 3D, and I'm sorry if I'm missing something. I don't mean to beat it up, but Sean, why did you withdraw the 3D Intelligence filing rather than just answer the agency's questions? I'm not sure if that's sort of simulator related, but I just didn't have my arms around that. Just to be perfectly clear, still with the new approach that you're taking, this is still going to have sort of the three options, if you would, in terms of, these are my words, but original, enhanced, and maybe importantly, the conservative option, if I've got that correct.
Sean Saint: Yeah. None of that has changed. That's a really important clarification on why we withdrew. The simple fact is that we got to the end of the 90-day review period, and some of these questions kind of came up near the end, and that's okay. That happens sometimes, and we just didn't have the time to get through them. When you get into a situation like that, at times you can get into a situation where the agency's like, "All right, look, I'm at the end of my time. I have to close this file." So I can either quote, disapprove it or you can pull it. It's just kind of a procedural thing. Everybody always does is pull it, answer the question, and resubmit.
Sean Saint: But we would've loved to have answered the questions within the bounds of the previous file, but we were just at the end of the time, so we didn't really have that opportunity, unfortunately. That is actually the point of why we say there is still a chance of getting clearance for this in Q4. Because again, if we pick up where we were, then potentially that's actually possible. Because again, look, we acknowledge it. It sounds bad. It's not as bad as it is. It was just literally we ran out of time to answer all the questions. It wasn't a problem, per se.
Jonathan Block: Understood. Thanks, guys.
Sean Saint: Actually, let me say one other thing on that. I'm not a MDUFA expert here, but my understanding is that the agency has a certain number of their files that they're supposed to close within their statutory 90 days, and then they're allowed to go beyond with a few. So we do see 510s that go beyond. It happens rarely, but it does happen. Most of them they're going to try and close. So that's just what happened here. But if people are like, "Oh, geez, I've seen them go beyond," that's absolutely right. But most of them do not. They try and meet their metrics on that. So we try and help them do that.
Operator: Thank you. Our next question comes from the line of Mathew Blackman from TD Cowen. Your question, please.
Mathew Blackman: Good afternoon, everybody. Can you hear me okay?
Sean Saint: Yeah, I got you, Matt. How you doing?
Mathew Blackman: Good. Doing well. Thanks, Sean.
Sean Saint: Yeah.
Mathew Blackman: Question for you. I was actually hoping for maybe a, let's call it a state of the union on the sales force, obviously ahead of the Mint rollout, but particularly given your comments about needing to drive awareness of Mint earlier in the call. Just anything you would be willing to share on just the, again, the state of the union on the sales force, how the new reps have been ramping? To the extent you want to talk, perhaps not, just about the strategy you have in terms of the Mint rollout, who you are going to target out of the gate, et cetera. Just again, any sort of updated thoughts on the sales force ahead of a pretty important commercial launch?
Sean Saint: Well, let me start with the awareness that we discussed earlier. The sales force, while they absolutely talk to users every day, primarily are calling on healthcare providers. What we weren't sure about was the likelihood of a user who's not seeing their healthcare provider, just sort of at home hearing about Mint in some way. Are they following us on socials? Are they checking out our website every day? Hard to predict how often that happens. But we don't think we have a problem getting the word out about Mint to healthcare providers, which is what our sales force is primarily engaged in doing. That was what the expansion was all about. It was about preparing us for that moment when we have Mint readiness, which we're nearly upon now. So sales force ramp.
Sean Saint: I would say that you can see it in the numbers at some level. It's absolutely sort of decelerated, and that's absolutely expected. We did a really good job, I think, of getting the word out on iLet with our previous, somewhat smaller group of reps, and we've had now three plus years to get the word out there. Getting more reps in the field doesn't really inform too many more healthcare providers about iLet that weren't informed about it before. However, you are going to start to slowly increase the rate of scripts, and that's exactly what we see.
Sean Saint: However, on the cusp of a major launch, as you alluded to, we want to make darn sure that we can get that word out to as many healthcare providers as possible quickly with the number of times you have to visit them for them to be comfortable, again, in a short period of time. So we really needed to do this expansion prior to the launch of Mint to do that. Hopefully, that's some additional color for you.
Mathew Blackman: I appreciate it. Thanks so much.
Operator: Thank you. Our next question comes from the line of Stephanie Elghazi from BofA. Your question, please.
Stephanie Elghazi: Hi. Thanks for taking the question. I wanted to ask on the process for getting Mint on formulary and how that factors into the revenue ramp for Mint next year. Is that process all factored into that 40% covered lives you mentioned for Mint at launch? Or is the, I guess, immediately addressable patient opportunity smaller because of that?
Stephen Feider: Hey, Stephanie. Yeah, great question. Look, you can hear it, I'm sure, in Sean and I's voice as we talk about the business that we are hugely excited for the upcoming Mint launch. Massive milestone for our company and allows us to compete in a very different way. If you are an investor or you are an analyst looking at the opportunity, I think you are naturally asking first, is there demand for the product? I hope we have convinced you, yes, by showing what the differentiation of it is. Then you are asking, can we manufacture the product? I think Sean went into a lot of detail today that illustrates that. But then you are getting to a really important point, which is how can you guys be sure that Mint will be covered, or the market access the insurance coverage on the product.
Stephen Feider: Again, reiterating that we expect 40% insurance coverage at launch, growing to at least around 50% by the end of 2027. Yeah, so we are in a good position to do that for a couple of reasons, to obtain that coverage, that 40% number at launch for a couple reasons, and very notable ones. Number one is that Mint is considered a patch pump for reimbursement purposes, meaning it is a Medicare Part D product, or at least that is the reimbursement designation. There is a nerd level term here called the GPI code for the product that governs whether the product is considered a disposable product. Mint is. On both Medi-Span and First Databank, we are considered that distinction.
Stephen Feider: By default, we are considered a patch pump, which makes the reimbursement process dramatically easier than taking a tubed pump, which is considered medically reimbursed and achieving pharmacy coverage for it, which is what we have done historically. But in this case, with Mint, it is a lot easier. So that is number one, it is considered a patch pump, which makes the reimbursement process much easier. Number two is we know how to do this. We have the capability. There are three large PBMs that represent over 80% of all lives in the country. There are 100 underlying health plans associated with PBMs that, combined, make up roughly 90% of all lives in the country. We have relationships with these decision-makers at these accounts. We know how to do this. We have a market access team.
Stephen Feider: We have laid all the groundwork associated with the Mint launch for reimbursement purposes with everything we have learned from iLet reimbursement. So I could go on and on about the individual accounts and details underlying this, but that is why I have confidence in these numbers we are promising here for Mint launch, and that is a very important item, going back to my first point, if you are buying into the success and the outcome associated with Mint.
Operator: Thank you. Our next question comes from the line of Richard Newitter from Truist Securities. Your question, please.
Richard Newitter: Hi. Excuse me. Thanks for taking the question. Just maybe going back to one that was asked earlier, I think lots of permutations as to where kind of the demand could eventually be sourced from for Mint. One of the things I was curious on is any quantification you could put around the existing iLet users in the pharmacy channel currently, and what you think that mix will look like as a percentage of the total Mint install base when we are maybe looking back at the end of 2027. Is it reasonable to think that your existing iLet conversions could represent, I do not know, what, 30%, 25% of whatever installed base of Mint users you have by the end of next year? I am just trying to get a sense for what is converting existing iLet in pharmacy channel users and what is truly incremental.
Stephen Feider: Mm-hmm. Yeah, Rich. Fair question. I am not going to give you the numeric answer for what I ultimately believe and I have modeled, but I will give you a directional sense of what I think will happen here. I do think that many iLet users, for reasons I have communicated earlier, are going to prefer to wear Mint, for reasons I hope that are clear to you all. I do also believe that much of our insurance coverage that we will have on Mint will come from the existing contracts that we have on iLet and added amendments or relationships and wins that we have already had associated with iLet when we are trying to achieve coverage for Mint. What does that ultimately look like in terms of what percentage of the new patients or patients we have next year on Mint?
Stephen Feider: Well, it is important that you know this first, is that we will likely be prioritizing to the extent where there is constraints on manufacturing capacity relative to demand. We will likely be prioritizing new users to Beta Bionics. At that point, I think certainly at our initial launch, where we are more likely to be capacity constrained, that will be resistant or that will be headwind towards more of our Mint install base being users who are coming from iLet.
Stephen Feider: But over the course of the year, again, because I think a lot of those iLet users who are getting their iLet reimbursed in pharmacy today, will have the ability to transition to Mint. I do think that it is plausible that we will see some of that conversion. I also think it is plausible that people who purchased iLet and are purchasing supplies through DME for iLet will also convert to Mint.
Stephen Feider: Now, important to remember there, that that is not a bad thing for us at all, financially. Patients who are purchasing iLet supplies from us in DME, we are generating directionally around $70 a month. Mint's reimbursement will be somewhere in the neighborhood of six times the amount of revenue per month per patient than what we are getting for that iLet patient in DME. So there is a little bit of a mouthful there without, again, giving you the direct number, but I hope that is helpful.
Richard Newitter: Yeah. No, it is. Maybe just another one. Just going back to churn and type 2, anything you are seeing in the market? You have been asked this question, I am sure, in the wake of Q2 results from a competitor of yours. Anything you have noticed on type 2 and churn retention that is worth calling out?
Stephen Feider: Yeah. We are not seeing any notable change in our churn or in our ability to retain patients than what we have over the last several quarters. Generally, retention on the product remains really strong, and I think you will see that by looking at our pharmacy. The best evidence of that is by looking at our pharmacy supply revenue relative to our installed base or our perceived installed base that you all have calculated. Our type 2 retention on the product is actually just slightly better than our type 1 retention. I am not advertising that as some grand difference between the two or some incredible characteristic of our business. I am just answering the question. There is no massive difference, but there is some discernible difference between type 2 and type 1 in terms of retention on our product, and type 2 is slightly better.
Stephen Feider: I have thoughts on why that could be, but just wanting to get you the direct answer.
Richard Newitter: Okay. Thank you.
Operator: Thank you. Our next question is a follow-up from the line of Stephanie Elghazi from BofA. Your question, please.
Stephanie Elghazi: Thank you. Sorry, I missed my follow-up earlier, but I just wanted to ask on the market, given there will be multiple patch or tubeless options in the market next year. Just curious how you are thinking about potential market expansion versus share taking. I guess, how do you think about your competitive positioning and anything you would be willing to say on your expected share capture potential on new patient starts with Mint? Thank you.
Sean Saint: Yeah. Stephanie, we are still in our infancy compared to some of the larger incumbents here. When we think about that question, we think primarily about designing the right product and whether we take share or whether we expand the market, we do better either way, and we do that by designing a better product in both cases. I am not trying to be obfuscatory here. The answer is both, of course. In the past, there have been people who have chosen patch pumps for good reasons. They like that format. Those are people that we are giving an additional option to today with a product, or not today, but soon, with a product that we believe has additional benefits.
Sean Saint: There are other people that have chosen not to move to patch pumps today or any pump today, because they did not find the product that best met the needs for something they were looking for. We believe that we are doing that as well. But I could not necessarily tell you exactly what portion of our expectation is from expansion versus share taking. For us, it is not really that important a distinction. Hopefully that is helpful. Actually, I will also say that I think it just illustrates the fact that what we have done is we have focused on making the best product that we can.
Sean Saint: I think everybody has heard us talk about this for the last year or so. Mint absolutely represents the right architecture as far as we are concerned. We think we have given users the right options. We think we have given them the right features, the right architecture. Is it 100% perfect for all users? No. But that is never going to be the case. What we think we have done is made a product that is most applicable to most users, which is exactly what Beta Bionics has always been engaged in, right?
Sean Saint: We are trying to improve the outcome for more people with diabetes, and you do that by making a product that more people will use, that the lion's share of people will use, and that is what we are doing. We think we have done a pretty good job of it, and we are really excited for 2027 to hopefully show the world that.
Stephen Feider: That's it for questions. Anything else you want to say in closing?
Sean Saint: Maybe just reiterating what I just said, which is 2027's going to be a fun year. We're really excited for Mint. I think you hear that in our voices. You hear that in the plans we're making. We do believe it's a great product. Look, I'm personally wearing it right now, and I'm loving it. I hope to bring that experience to a hell of a lot more people. Thanks to everybody for joining today. We really appreciate your time, and I look forward to catching up with each one of you.
Stephen Feider: Take care, everybody.
Operator: Thank you, ladies and gentlemen, for your participation in today's conference. This does conclude the program. You may now disconnect. Good day.
