Genscript Biotech Crp Ord

Good day and thank you for standing by. Welcome to Genscript Biotech 2026 Interim Results 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. To ask a question during the session, you need to press star 1 1 on your telephone. You will then hear an automated message advising your hand is raised. To withdraw your question, please press star 1 1 again. Please be advised that today's conference is being recorded. I would now like to hand the conference over to your first speaker today, CFO of Jamescript Biotech Group, Mr. Phil Zhao.

Please go ahead.

Welcome to our 2026 Interim Results Conference call. Joining me on the call today are Mr. Robin Munn, Chairman of the Board, Mr. Sherry Shao, Locating CEO of Genescript, Dr. Rui Chen, President of Genescript Life Science Group, Dr. Icy Bai, General Manager of BestSign, and Mr. Alan Wolk, CEO of Propel. During today's call, we will be making statements about future expectations, plans, and prospects, as well as any other statements regarding matters that are not historical facts, which may constitute overlooking statements. Actual results may differ materially from those indicated by such overlooking statements because of various important risk factors and changing marketing conditions. We do not undertake any obligation to publicly update any overlooking statements. Before we begin, please note that the prior period of figures presented in this conference hall are on a comparable basis, excluding the financial impact of the license transaction with Lenovo pursuant to the license agreement that was recognized in the prior period. We believe this provides a more objective view of the group's underlying business performance. Today, Sherry will provide an overview of our company performance and growth drivers. I will then guide you through the financial performance. Following that, Sherry will update on our full year guidance. We will also have a Q&A session at the end of the call. As a reminder, today's presentation and recordings will be available in the investor relations section of the company's website. Now, I will hand it over to Sherry.

Thank you, Phil. Before turning to company performance, I'd like to provide an overview of our company and the co-drivers behind our long-term growth. Jamstretch is a leading global platform for lifetime services and products. We serve more than 260,000 customers across over 100 countries and regions. Supported by an integrated R&D, manufacturing, and commercial network across North America, Europe, and Asia-Pacific. In the first half of 2016, James Quirk delivered a strong set of results that demonstrated both growth momentum and improving business quality.

Samuel exceeded $400 million, with

37.3% year-over-year growth, while adjusted net profit reached $62.5 million, growing more than 200% year-over-year. More importantly, these results reflect the early benefits of a more scalable growth model. AI-driven demand is Expanding our addressable opportunities. Our Gen2Protein platform is deepening customer value and competitive mode.

And operating leverage is translating growth into stronger profitability.

Now, let me walk you through key operational highlights across our business for the first half. Biotech Crp Ord Biotech Crp Ord Biotech Crp Ord Biotech Crp Ord Operational Efficiency and Productivity at Scale Importantly, our Gene2Protein platform continues to change the game by compressing turnaround from digital sequence to model-ready data to as fast as 4-phase, the fastest turnaround in the industry. This is precisely the kind of speed the AI heroes demand. Provider maintains daily growth with accelerating order momentum. Increasing cap line moving to downstream driven by AI and in-level CAR T is becoming a new growth engine. While improving scale continues to lead to profit village. For Bethlehem, rising custom adoption continues to validate the commercial value of our products. We are also applying AI to protein design and engineering to accelerate innovations and their launch in the market with impact.

Our growing IT portfolio underpins our long-term successes. As we look ahead, AI is revolutionary in drug discovery.

I will highlight how these disruptive trends in AI drug discovery are unfolding and why Janscoot is perfectly positioned to leverage this shift and drive this revolution forward. AIDD is approaching a critical tipping point. As summarized by Anthropix briefing, Over the past few years, AI has fundamentally reshifted the software development industry, and life sciences is emerging as the next high-value application beyond software. We see increasing AI applications in target delivery, molecular design, and candidate selection. According to another survey by Deloitte,

Nearly 60% of surveyed biopharma leaders ranked AI in research and discovery as a top priority. Indeed, the potential of AI in drug discovery is undeniable.

First, AI has the potential to dramatically accelerate early-stage discovery Compressing a traditional 4-6 year timeline into just 12-18 months 7. AI can enable researchers to explore a much broader range of therapeutic targets, modalities, and political designs at computational scale 3. By identifying high-quality, winning candidates early before more costly downstream development begins. Automatically, AIDD models promise higher success rates and a maximized ROI.

Here is the structural shift we are tracking.

Applying AI to the licenses and drug discovery is fundamentally more complex than pure tech. Biological data is inherently less complicated and our feedback loops, demands, resource, physical web lab experiments. But make no mistake, the AI design transformation of our industry is inevitable. Those who can bridge the gap between digital design and web lab reality

will dominate the market.

The key challenge facing AI-driven job discovery today is no longer generating ideas and designs. It is validating stuff. AI models can now generate thousands of candidate molecules in hours. However, traditional experimental workflows were not designed for that scale of speed. Validation often takes weeks, involves multiple disconnected steps, and produces data that is not always ready to feed directly back into AI models. As AI accelerates design, the bottleneck is shifting decisively toward Experimental Validation We are experiencing a significant opportunity emerging around a new category of infrastructure. Validation platforms that are fast enough for AI iteration, scalable enough for AI volume, and structured for continuous model learning. That is where transcript is uniquely positioned.

We are here to close the loop.

To address this industry challenge, we put toward our four-day AI to biology validation engine. This is an integrated validation platform built specifically for the AI and it will continue to improve and evolve. Our platform combines four critical capabilities. First, scalable capacity. Our modular gene-to-protein and assay workstations allow us to drastically add throughput as customer demand grows. Second, speed. We can move from digital sequences to model-ready biological data in as short as four days, dramatically reducing validation timelines. This is industry leading, world leading. Third, integrated execution, automated and digitally-attracted workflows, reduce manual handovers.

Third, improved consistency and

Skill Efficiency For AI-ready data, results are generated in formats that can support model iteration and continuous learning. Together, these capabilities flow transcripts to help close the gap between digital intelligence and biological execution, transforming validation from an industry bottleneck into a next-generation solution and our competitive advantage.

Here's why this matters commercially.

This shift isn't just technological. It's redefining our customer base and our revenue model. AI is changing who buys from us and how they buy. Beyond our co-pharma and biotech base, we are now increasingly engaging AI-native biotechs, model developers, and major technology companies entering life sciences for the first time. A generally new commercial category for life sciences, industry, and for gen-script.

AI DE is changing how our customers operate.

The AI models generate more designs from more experiments and iterate faster, which turns validation from a one-off and longer cycle project into a recurring, ongoing stream of demand. That's it. The new segment of customers and evolving operating models from our customers is what makes this opportunity structurally larger than a typical cycle. And the result is report a larger customer universe, higher order volume, faster cycles, and deeper, longer term engagement. Together, They create a larger and more strategic growth opportunity for transcripts over time.

For investors who are newer to the AI for Science ecosystem, this slide illustrates where transcripts sit in the value chain.

AI models can generate designs, but those designs must ultimately be translated into biological contracts, validated experimentally, developed into candidates, and ultimately manufactured at scale. Genscript participates across multiple points of that workflow. Our life science group leads validation through our ginger protein and acid platform at an unmatchable speed and skill. We are closely integrated into the AIDB group. The BIOS helps advance promising candidates into development and manufacturing once pipeline advances. SM leverages innovation and AI-enabled protein engineering to create new opportunities in Synthetic Biology. Combined with our global operating footprint and strong balance sheet, these capabilities position GenScript as a critical infrastructure provider, supporting the next generation of AI-enabled biotech innovation. We are not simply advocating in AI for science ecosystem. We are actively building the validation infrastructure that enables it. As AI-driven discovery scales, transcripts will continue to strengthen its position as one of the most strategic and valuable points in the biotechnology value chain. Having discussed the strategic drivers shaping our Long-Term Opportunity I will now hand the call over to you to review our finance performance in greater detail.

Thank you, Shari. Let me now take you through the group's financial performance for the first half. RealMail reached $404.2 million, up 27.3% year-over-year, reflecting strong momentum across the group. First was FrostBase. Last time, services and products grew 28.8% to $390 million, profile grew 34.2% to $61.1 million, and the baseline grew 7.4% to $30.4 million. Alongside the top-line growth, we also delivered higher-quality earnings. Group growth profit reached $206.7 million, up 48% year-over-year. Significantly outpacing well-known growth. This reflects our intrusive business mix, operational efficiency, and scale benefits. Benefiting from well-known growth and intrusive operating average, a drug is net profit reached $62.5 million, up 203.3% year-over-year. A report for AnyHub and a clear guidance, clear evidence that profitability is scaling faster than the top line. Overall, the group delivered growth across revenue, gross profit, and net profit, which is a strong validation of our ability to create long-term value by leveraging our global footprint, innovative platforms, and skill. Now, let's turn to the Lifescience Group, or LSG. In the first half, LSG delivered strong revenue growth alongside meaningful profit expansion and operational efficiency gains. Revenue reached $319 million of 28.8% over the year, around 10 percentage points above initial guidance. Growth was driven by sustained global customer demand, increased penetration of gene to protein platform, and rapid expansion in AIDD-related demand. We are seeing strong demand for high-quality gene synthesis, protein expression, and related research services. across pharma and biotech customers and AI-driven companies. More importantly, profitability improved significantly. Adjusted gross profit reached $185 million, up 46.1% year-over-year. Adjusted operating profit reached $94 million, up 102.8%, surpassing $90 million for the first half and effectively doubling. This benefits from our improved operating leverage. Over the past couple of years, we have consistently invested in automation, digital operations, capacity expansion, and our global performance. Alongside business growth, we see higher operational efficiency and a stronger profitability, enabling profit growth to outpace round-the-clock growth. Expanse trends were also encouraging. Growth in selling, administrative, and R&D expenses remained below round growth, reflecting strengthening scale effects and disciplined resource allocation. We also see improved margins. In the first half, adjusted growth margin reached 57.8% or 55.4%, excluding the impact of U.S. tariff refunds, and adjusted operating margin reached 29.5%, The operating margin approaching 30% marks an important milestone for LSG, transitioning from investment to growth at scale and profitability. With growing demand from AI-driven drug discovery and expanding global customer base and increasing platform synergies, we believe LSG is well-positioned to drive both realm growth and profitability improvement. This slide shows why LSG's growth is not dependent on single product, region, or customer type. LSG's sustainable growth is attributable to its leading platforms, global reach, and a broad customer base. Looking first at our product mix, Ginterplating products and services contributed around two-thirds of LC Rounding, making it our most important business area. This reflects both our leadership of a general pricing platform and a strong customer demand for integrated R&D solutions. By region, our Rounding base remains well-balanced. North America contributed approximately 50% of Rounding, while Asia Pacific and Europe accounted for 29% and 21%, respectively. This diversified global footprint allows us to capture opportunities across major markets while enhancing our business resilience. Our revenue stream is highly resilient, built on a strategically diversified customer base. With over 80% of revenue generated by pharma and biotech, we are deeply embedded in leading R&D engines. Complementing this, Our robust presence across global research institutions ensures long-term structural collaboration well beyond our industry segments. Taken together, our GMPTEAM platform, global operating network, and the diversified customer base provides a strong foundation for LSG's continued growth, enabling us to capture opportunities arising from AI-driven life science innovation. Moving on to the opportunities ahead and the key drivers that will support long-term growth, we see three engines powering LSG's growth in the years ahead. First, our integrated general pricing platform remains the primary engine of our growth. We are tracking the structural shift as customers transition from transactional, single-product purchases to our comprehensive end-to-end solutions. Biotech Crp Ord Biotech Crp Ord Biotech Crp Ord Biotech Crp For GenScript, this translates directly into significant larger contract values and exceptional long-term value visibility. We expect this momentum to compound aggressively with AIDD orders projected to double in the second half and maintain that hyper growth trajectory over the next several years. Third, we are seeing stronger returns from our platform investments. Biotech Crp Ord Biotech Crp Ord Biotech Crp will directly drive margin expansion and a superior shareholder value. Overall, the continued expansion of the gentle protein platform wrapped the growth in AIDD-driven demand and improving returns on our platform investments underpin LSG's high-quality growth over the next several years. Turning to ProBio, the business continued its strong momentum in the first half Delivering revenue growth includes profitability and greater operational efficiency under our end-to-end CRDMO platform strategy. Please note that all the year-over-year growth rates presented here are on a comparable basis, excluding the financial impact of the Lenovo license transaction. Revenue reached $61.1 million US dollars, up 34.2% year-over-year. continuing the healthy trend of recent quarters, driven by faster order execution, new customer wins, and progress across existing programs. More importantly, flat growth is now translating into profitability. Adjusted gross profit reached $8.3 million, up substantially from around $3.7 million in first half 2025. As part of the project mix, higher utilization and manufacturing efficiency all came through. Expense growth remained well below revenue growth. We kept investing in R&D and other technology platforms while tightening organizational efficiency, and as revenue scaled, fixed cost absorbed more effectively. Operating leverage is now clearly visible. This shows up most clearly in adjusted EBITDA, where the loss narrowed to $6.5 million from $16.8 million in first half 2025. An improvement of over $10 million is a meaningful step toward profitability. With investments made in platforms, global expansion and capacity are now converting into profitability as revenue grows. Looking ahead, we will stay disciplined on high-quality growth, driving revenue, improving operating leverage, and reinforcing Provel as a leading global CRDMO partner. Beyond revenue and margin, we are focused on the quality and sustainability of future growth. And on that front, our order intake stood out. On revenue, Provel grew 34.2% organically, Biologics Business grew 44.2% and Advanced Therapy Business grew 16.3%, brought base strength across both lines. The real headline is orders. New orders grew 54% year-over-year, significantly outpacing revenue growth. Biologics Business up 62.1% and Advanced Therapy Business up 34.9%. Our backlog continues to build, further enhancing the visibility of our future revenue. By region, we achieved steady growth across all major markets. On revenue, China grew 43.1% and international markets grew 30.3%. On orders, China grew 73.8% and international markets grew 55.3%, demonstrating robust demand across both markets, and solid-based speedy outcomes. Overall, Propel is delivering strong growth across round-new, new orders and market expansion. In particular, orders consistently outpacing round-new reflects customer recognition of our entry and CRDMO platform and reinforces our confidence in growth outlook ahead. Finally, turning to BestLine. Despite market headwinds, BestLine maintains steady growth while continuing to invest in innovation and commercial execution. Round will reach a $30.4 million US dollars, up 7.4% year-over-year, driven by rising demand for core products and growing customer base. Our expertise in industrial design and biomanufacturing continues to reinforce our competitive position. Profitability also improves. Adjusted Gross Profit Group 14% Year-over-Year to $13,000,000, outpacing round growth on better product mix and improved manufacturing efficiency. Innovation remains our core driver, with adjusted R&D investment reaching $5.6 million in the first half, spanning industry enzymes, biomanufacturing, and synthetic biology, while we apply AI and digital tools to improve R&D productivity and Speed Commercialization. Alongside that, we continued strengthening our commercial capabilities and global reach, expanding customer reach as demand grows for high-performance enzyme products and sustainable solutions. On the bottom line, adjusted operating loss was $1.3 million, compared to a $0.6 million loss in first half 2025. A deliberate investment in platform and innovation led positions to unlock larger growth ahead. Looking forward, with new product commercialization, continued market expansion, and emerging scale benefits, we expect Dashline to lift both revenue and profitability.

To conclude, let me share our outlook for the four years. Looking ahead, We are actively capitalizing the industry, helping the acceleration of AI-driven drug discovery revolution, the fast expansion in global biopharma R&D, and the next-generation transformation of global biomanufacturing. In Life Science Group, our mandate is clear. We will scale our leading ginger clothing platform, pushing segments to data delivery, with unmatchable speed and capacity. We will keep visualizing and automating our global network to improve efficiency and reliability. More importantly, we will be directly integrating our one-lap validating engine into our customers' R&D systems and digital infrastructure to power the future of AI self-discovery. In ProBio, we will continue to benefit from growing biologic demand and emerging opportunities in InVivo CAR-T and AIDD. We will expand our global footprint, strengthen our platform, advance more programs from early discovery into clinical and commercial stage, and stay on track for positive EBITDA in 2027. In fact then, we will focus on commercializing sweet proteins, accelerating AI-enabled RMD and product optimization, and continuing to expand globally while strengthening our IP position. Supported by our strong first-half performance and confidence in the opportunities ahead, We are raising our four-year guidance for the Lifeline service segment. We now expect revenue growth of 25% to 30%, adjusted growth margin above 55%, and adjusted operating margin above 25%. For providers, we are increasing our revenue growth guidance to 25% to 30% and continue to accept the business to achieve top-tier EBITDA in 2027. For Bethlehem, we expect revenue growth of 8% to 10%, while maintaining an adjusted growth margin of over 43%. Taken together, our platform leadership and continued investment in global bridge and innovation positions GenSwift to deliver high-quality growth and long-term shareholder value.

That concludes today's presentation. Operator, please open the floor for questions.

Thank you. As a reminder, to ask a question on your telephone, please press star 11 and wait for your name to be announced. Just a moment for our first question, please. First question comes from the line of Yang Huang from JP Morgan. Please go ahead.

Thanks for the opportunity to ask questions. I have two questions. Our first one, then a follow-up.

So we noticed a significant

Biotech Crp Ord Biotech Crp Ord Biotech Crp Ord Biotech Crp Ord Biotech Crp Ord Biotech Crp Giving AIDD demand and AIDD demands remain very strong, how should we think about the likelihood of further upside to the current life science guidance? Are you seeing any signs that such demands from AIDD will continue to outpace your existing assumptions? Thank you. That's the first one.

Thank you, Yang, for your questions. I'm happy to answer. This is Ray from NTSC Life Science Group. And according to your questions, let me answer in this way. There were three things that drove our growth, and they reinforce each other rather than standing alone. First, the AIDD demand itself has more AI-related biotech foundation model developers and innovation-focused pharma groups. Biotech Crp Ord Biotech Crp Ord Our customer base is compounding and getting higher quality. We're winning new AI-focused accounts while maintaining the healthy demand from our traditional pharma and biotech base. More importantly, our growing share of customers are converting from single-project work into multi-stage, multi-service partnerships, which is what turns one-time projects into recurring revenue as partnerships. And third, We are now reaping the fruits of our years of platform investment. Sustained spending on automation, high throughput capacity building, and digitalization is translating into shorter turnaround, higher utilization, and greater scalability. And that operational leverage has become foundational to how fast we can further grow. And for the second part of your question, we do believe that ADB is still early and customers are moving from proof of concept into large scale validation and iterations and optimization. And that's exactly the phase where demand for high throughput gene deproting productions and high quality data accelerates. We're seeing that in three concrete signals. The sizes of the order is growing. The customer's relationships are deepening. And the share of a high complexity and structured recurring orders, an area that we are uniquely strong, is rising. And that last point matters most because it's what gives us confidence and visibility for long term rather than just momentum. Wanted to be clear, this growth isn't the result of one larger customer or one project. It reflects a structural shift in demand and our unique ability to convert that demand into revenue and profitability at scale. Looking ahead, we are continuing to manage guidance prudently, however, and the underlying demand signals, project volume, size of the order, customer deaths, and the growing mix of complex recurring ADD projects are all pointing the same direction. And big farmers are evolving and also adopting as well. So it's a fascinating time in the industry. If these trends continue and our unprecedented execution remains strong, and we're confident about that, too, and we will see a real potential for continued upside versus our current assumptions.

Thank you. Okay, great. Yeah, great. Thanks for the answer. My second one is competitor landscape. To our understanding, TraceBioScience is one of our, you know, company's primary competitors in the gene synthesis market. So how does management view the competitive landscape between Twist and the Genescript in the AIDD space related to business?

Thank you, Yang, for your question. Again, this is Ray. I would like to be a little bit more specific here because we think the data speaks for itself. At Genescript, we don't compete on commodity volume. where we compete on value per delivered results, the speed to data, and the data quality that customers can actually rely on. Let me give you some numbers behind that. First, on economics, Descript captured more than twice the revenue per delivered item versus the company you mentioned, and that gap has continued to widen. On throughput, We're processing in real more than 4,000 designs per day, meaningfully ahead of the publicly reported numbers from the company you just mentioned, which is in thousands per week. The third about speed. We deliver from physical sequence to binding data in 47 calendar days, depending on which expression route the customer is taking. Comparing to more than two weeks reported error somewhere in the market from the company you mentioned. The number is important here because this is three to five times faster iteration cycle, which matters enormously for customers who is doing AI model. Depends on the continuous experimental feedback in the loop. We think the most important edge actually is the data quality. In one recent customer round evaluation, our assay variability came in 110% compared to close to 30% for the company that you mentioned in their workflow. So in the AIDD area, reliable, low variable data is not a must to have. and that's where we believe our uniquely strong and clear advantage is most durable. And I can explain a little bit more about our fundamental business model differences as well by comparing the company you mentioned. Much of the competitive landscape especially, they stop just at DNA or fragments. Biotech Crp Ord Biotech Crp Ord Biotech Crp Ord Biotech Crp Ord Biotech Crp Ord and this allows us to add capacity faster with minimally lower capital intensity. Eventually, we're targeting doubling the throughput of our capacity every quarter, and we have the competence to do that, and we have to sustain in the industry meetings pace. That's what we have been committed to. So, looking ahead, our objective is crystal clear. serving the customers to be the definitive and always industry-leading biology validation engine, especially for the AI drug discovery era. And years for the speed, the scale, and the reliability the industry now demands and requires. And we need to build to stay there. Thank you, Yang, for your question and allow me to have the opportunity Thank you.

Okay. Thank you, Ray, for the detailed explanation. Yeah, I finished my questions. Thank you.

Thank you very much. Just a moment for our next question, please. Next, we have David Chong from Jefferies. Please go ahead. Hi.

Can you hear me? Yeah. Thanks for taking my question. My first one is about your profile. We noticed that profile has began circling AI drug discovery projects in the first half of 26. Could you please briefly discuss the current pipeline of drug discovery related orders and profile technology capability is about We found both lifestyles and the CIDMRO has delivered a meaningful acceleration in the growth. To the management, elaborate on this capital allocation and tax plans to support this opportunity, and also to investors, we expect any incremental financing requirements, all fundraising activities as company continues to to expand its capacity and capability.

Thank you, David. So, for the first question, Allen will help address, and for the second one, I'm happy to address the questions.

Thank you. Thank you for your question, David. This is Allen of Provile. So, for the first half of 2006, Provile totally signed a $9.7 million AIDD-related order, including both for discovery and also CMC project. And actually we delivered around 2.5 million revenue. So the AI generated drug candidates present a unique development requirement and increasingly demand integrated solutions across discovery and also development. Based on our experience with the AIDD program to date, we have several common characteristics. Drug candidates guaranteed or optimized by AR are typically complex molecules, including bispecific or trispecific antibodies, which require further validation and optimization to address developability and drugability conservation. Second, the targeted selection and therapeutic applications are becoming increasingly diverse, spanning multiple disease areas and modalities. Third, customers typically require significantly accelerated development timelines to maximize the efficiency advantages delivered by AI-driven discovery. And fourth, some customers will advance multiple candidate molecules simultaneously, and this creates demand for high-throughput and also parallel development capabilities. To address this need from AIDD, Tobai has quickly established a specialized solution across both the discovery and also CDMO value chain, based on our extensive experiences, know-hows, and platforms. First, leveraging our more than 20 years experience in the biologics discovery segment, for biologics discovery provides a comprehensive rapid lab validation platform designed specifically for NIDD program. The key capabilities include customized data generation, and experimental support for AI model training and optimization. Second, highly automated and high-throughput workflows capable of processing thousands of samples a day. And third, integrated in vitro and in vivo pharmacology platform across multiple assay formats. And fourth, tailored screening strategies that optimize both biological activity and the availability. So this end-to-end approach Provide can help customers advance AI-generated candidate to PCC within as little as four months. And secondly, for our biologic CMC platform, we have introduced a combined developability assessment and extract CMC offering tailored for AI-derived molecules. The key advantages include early stage developability assessment, with leveraging the same host cell system and expression vectors employed in downstream CMC development, and rapid identification of sequence liabilities and potential CMC challenges before entering formal development, meaningful reduction of downstream technical and manufacturing risks. And we expanded our accelerated development framework beyond monoclonal antibodies to include isometric, bispecific, trispecific antibodies, and also high-concentration programs. For even the most challenging molecule format, for biotin programs from cell transduction to toxicology batch production at later as 4.5 months, significantly shortening the development timeline for AI-driven programs. AI-DD is really moving incredibly fast with the profiles, Yeah, thank you, Alan. So, David, for your question regarding our capital allocation and PEPX plans, so, yeah, we will keep investing into match these growth not just the chases.

Expanding capacity, automation, and technology platforms to support the robust growth of both the LCG and CRDMO business. Our approach to capital allocation stays very disciplined throughout, with every investment calibrated tightly to customer demand and expected return. So let me give you some numbers. So during the first half of 2026, the group incurred a capital expenditure of US$48.4 million. And based on current basis of momentum and the project execution progress, we expect the year 2026 capex to remain at a healthy and flexible level with total capex approximately US$130 million. Importantly, the company continues to maintain a robust balance sheet and a healthy liquidity structure. We currently hold approximately US$830 million in cash and cash occurrence. In addition, our business and management of capital expenditure, operating costs, and the return on invested capital, ROIC, has driven a significant year-over-year improvement in both working capital and free cash flow during the first half of 1026. So, given our substantial cash position, our continued ability to generate operating cash flow and high-quality enterprise credit standing, we are confident that we have sufficient financial resources to support all ongoing expansion initiatives. And therefore, we do not view growth and investment as competing priorities. Our objective is to continue investing aggressively in the highest return opportunities while maintaining discipline in the capital allocation and creating long-term shareholder value. So going forward, we will keep advancing our growth strategy with the same financial discipline and that got us here. So in positive space, we are doing so. from a position of increasing profitability, including return on capital and growing financial strength. So to directly address your second question, we have a sufficient liquidity buffer, so we see no need for equity financing.

Thank you, David. Thank you. Thank you.

Thank you. Just a moment for our next question, please. Next, we have Laurent Tam from Morgan Stanley. Please go ahead.

Laurent Tam Thank you. First of all, congrats to management on these fantastic results. I have two questions. The first one is that within the AIDD customer base, we have noticed that large AI model companies are gradually becoming a new source of revenue growth. Could you help us better understand the profile of these customers? and how their needs differ from those of traditional pharma and biotech companies. That's my first question.

Thank you, Lawrence. This is Ray. I'm from Chesco's Life Science Group again. It's a really, really interesting question, and we're learning as well along the way. This is one of the most important shifts we're seeing in our customer base. and it's worth being precise about what we have learned and how differently these customers behave. Let me start about the scale. A traditional pharma program typically advance a handful of candidates through a small number of projects, but the AI-native customers operate on an entirely different order of magnitude. and their models can generate hundreds, thousands of candidate sequences in a single iteration. And each of those needs rapid experimental validation. That alone changes the order size minimally. And second, about speed, it's truly different. A traditional design test cycle runs in months, and AI-led customers need Continuous Fast Feedback Loop because the model requires a steady stream of experimental data to retrain and improve, which is why turnaround measures in base now and not weeks. And high-throughput expression, automated platforms, it's must-have for those customers. And third, it's about the molecular complexity, and it's different. The traditional pharma projects concentrate heavily on well-validated targets or formats. But AI-native customers, they are pushing the boundaries. They have more complex designs because the models are explicitly designed to explore molecular space that traditional discovery won't even attempt. Here's the insights that matters most, we think, how we will further grow. For AI-nated customers, we're not just delivering a DNA construct or protein itself. We are delivering the data. The data is the end goal. And the expression, functional activity, binding, stability, developability data, all feedback into the model development as input. This means we're not just selling one project or a deliverable. We're becoming part of the customer's AI development infrastructure, which is truly, truly fascinating. And that's what's driving the real aspect, so that the customer's lifetime value. The traditional engagement is often the one-off project, and with AI-related relationship can expand the complete workflow in drug discovery. which can also further extend to the clinical and CMC like Alan just mentioned with ProBio. This is always one partner. So simply put, AI-lative customers that demand more scale, more speed, more molecules, more designs, more data earlier in the process than traditional programs ever did and accomplish Biotech Crp Ord Biotech Crp Ord Biotech Crp Ord which is more exciting as well. The traditional pharma is evolving as well. They are adopting together along the way. So this is why we see that the customer segment is a structural, a long-term growth driver rather than a short-term momentum. And thank you for allowing me to share what we have learned.

Thank you, Ray. My second question is on BestSign. Thank you, Lawrence, for the question. This is Aishi from BedSign.

So AI has become deeply integrated into every state of R&D at the best time. And the business impact comes down to two things, larger R&D and lower cost. So first, our trend and function models let us optimize enzyme performance across multiple dimensions. The enzyme molecules we are getting today perform at levels that our old methods simply could ever reach. Second, our AI-driven multi-property optimization has doubled our positive hit rate compared to 2025. In the best cases, we can hit our project goals in just two rounds of wide range design with fewer than 150 newtons. This not only improves our project's success rate but also significantly increases the number of projects we deliver. For 2026, we expect to deliver six to seven projects, which is twice as many as in 2025. So, we will start development cycles. We will launch both the design online platform and the product flow agent platform. They are already cutting on the time by 20%, and on top of that, the agent platform lets more of our scientists design and optimize enzymes simply by using natural language. So it's much more accessible to the team. Our protein AI models help boost enzyme activity, while our DNA-related models help improve our production yields. So put together, they drive significant cost reduction. Over the past year or two, we've seen meaningful cost savings across more than three projects. Take our 5E and high-temperature AMLEs as examples. Both achieved better performance at a lower cost with gross margins up by 7% and 6%, respectively. for the single product.

Thank you, Laurie. Thank you.

Just a minute for next question, please. Next, we have Linhai Zhao from Goldman Sachs. Please go ahead.

Thanks for taking my question. Congrats on the great results for Life Science Group in particular. I'm interested to get more color on the improved gross profit margin in the first half. It seems like the AIDD orders came with a higher margin compared to the traditional change of protein orders. Can management share more colors on that? And given that, can we get a better sense moving forward, how should we think about the long-term AIDD margins? if the ARD margins would likely to remain higher? How should we think about the entry barriers into this field and the competitive mode of transcript? And also, Dr. Chen also mentioned that increasing utilization to cope with the demand as we're expecting triple digit growth in the second half and even beyond. What is the utilization rate that we're currently seeing and how are we preparing for the increased capacity going forward, especially for protein size? Well, I understand that the automation level is lower compared to the gene part. Thanks.

Thank you, Linhai. I will address your first question regarding the profitability of AIDD orders. And for the second part, visualization, I will defer to Ray to answer. Okay. So AIDD related orders carry structurally higher volume than traditional protein extraction version. Please don't worry about explaining exactly why, because that's why these are durable, not one time. First, customers typically face a large-scale, high-volume orders, which significantly increase protein expression throughput per project, and that's a portion of the fixed cost across our platform. Second, they normally pay premium for seed. Drag it around isn't an option for them, because their AI models depend on Faster experimental feedback to keep iterating and see that the scale is exactly what our automated high-throughput infrastructure is used to deliver. Third, and most important, the deliverable is the item itself. It's different. Customers aren't just buying a protein. They are buying high-quality, structured, experimental data that feeds directly back into their models. Biotech Crp Ord Biotech Crp Ord Biotech Crp Ord Biotech Crp We believe this differential holds and could widen for two reasons. One, the value we are pricing for it. Let it be the scale and the model-ready data, as Ray just mentioned. You know, it becomes more variable and SAI adoption in drug discovery defense. And because of the cost of validation bottlenecks, holding grows, you know, more painful. for customers at their model scale. And second, as our platform utilization increases and automation investments mature, our own cost structure improves in parallel, which means we can defend this margin premium even as the category grows and the competitive intensity increases. So that is a foundational and structural advantage tied to how all this is built and built to last.

Thank you for your question. I can talk with you for days about how we could scale. The throughput is not only the genes, but also all the way to protein expression and the first downstream average for the validations. So our throughput, our Infrastructure is built as I mentioned the modular and the intelligence workflows and workstations and which allows us to scale very rapidly with confidence. And the orders, the magnitude that we're getting, no one else in the world could accept and deliver. That's what we're doing right now. We have the confidence to further double and doubling our throughput capacities in the coming days and coming months in a very exciting way. Thank you.

Thank you. Thank you very much. Yeah, thank you. And for your interest and the questions and ongoing support for GenScript, we apologize for not being able to address all the questions due to timing limitations So if you have additional questions, do not hesitate to reach out to our investor relations team. And we will see you on our next call. Thank you.

This concludes today's conference call. Thank you for participating. You may now disconnect.