Blaize Holdings, Inc. Q4 FY2025 earnings call

• In 2025, revenue showed strong growth, exceeding revenue guidance. There was approximately 20 times growth over the year across infrastructure, sovereign AI, and public safety applications.
• Strengthened execution by bringing on a chief revenue officer to scale commercial efforts globally.
• Signed an MOU with Nokia's Asia Pacific Division and is advancing collaboration through an innovation hub in Singapore, with early traction in the Asia Pacific region across multiple applications.
• In India, signed an MOU with the government of Telangana supporting its AI cloud innovation hub. In China, expanding footprint with regional solution providers. In Korea, working with solution partners like GSIL. Across Southeast Asia and Australia, working with Nokia and vertical systems integrators on various AI use cases. In the US, Europe, and Latin America, expanding engagements in multiple environments.
• Blaze model is cash flow efficient with lower silicon cost and power efficiency. A hybrid configuration can bring about 50% lower infrastructure cost and 60% lower power consumption.
• Progressing toward initial release of Blaze AI Services platform in the second quarter, combining infant silicon, intelligent software, API - based AI services, enabling customers to move faster from infrastructure to real - world outcomes.
• Strengthening partnerships with systems integrators and software providers to accelerate and streamline adoption of AI solutions powered by Blaze hardware and software.

In 2025, revenue grew from approximately $1 million in the first quarter to $23.8 million in the fourth quarter, exceeding the upper end of revenue guidance, representing approximately 20 times growth over the year. Full - year 2025 revenue was $38.6 million, outperforming the upper end of guidance by $600,000. Fourth - quarter revenue doubled from the prior quarter to $23.8 million, exceeding the upper end of guidance by $700,000. Gross margin for the fourth quarter was 11% and 16% for the full year. Blaze hardware and software is expected to form a higher mix in AI solutions from the second half of 2026, resulting in gross margins of between 30% and 35% in the fourth quarter.

• 2026 fiscal revenue guidance is $130 million, unchanged. Expect the first half to be lighter than the second.
• Flat gross margins for the first half of 2026, expected to average between 30% and 35% by the fourth quarter.
• Adjusted EBITDA loss of between $45 million and $50 million.

• Global memory supply constraints.
• Geopolitical tensions.

Q: When you talk in your press release about different types of applications, how would you prioritize them in terms of what you're going to have this year and how those opportunities play out over the next three or four years?

A: The commonality is AI inference. Near - term priority is converting the pipeline where we've got access to those customers and data. Medium - term is expanding more business into some of those customers.

Q: About gross margins, longer term and at scale, what do we expect our long - term model to look for in terms of gross margins on the hardware side and on the software side?

A: In the longer term, it's 55% plus, a blend of hardware and software. Software and recurring revenue could become a larger portion of the mix but too early to say for sure.

Q: For calendar 26, can you help us understand the extent to which Starshine and YADA are driving that versus other things like maybe converting the Nokia MOU into revenue or maybe even getting traction on some of the new capabilities that we identified in the press release that you've talked about the services platform and AI application delivery?

A: Yoza and Starshine are important partnerships, but the pace is driven by their end user needs. Other partners have been added and more may be added in the next three to six months. The AI services platform and relationship with Nokia are expected to feature towards the end of next year. Revenue falls into three big buckets: system revenue, attach rate of Blaze software, and API revenue from the AI services platform.

Q: Can you give an update on whether the opportunity pipeline of about $725 million is still the right way to look at it or has it changed? And on the adjusted EBITDA guidance for the year, can you clarify to the extent to which mask set costs are included?

A: Pipeline is dynamic, with meaningful traction in Asia - Pacific. Geopolitical tensions have impacted some deployments. The Nokia partnership and AI services platform will add to the pipeline. Mask set costs for the core chip design are in 2027 and beyond, with early third - party external costs kicking in towards the second half.

Q: Regarding calendar 26, how much of the $130 million guidance is in backlog or some sort of commitments here and how do we think about customer concentration or mix this year within that?

A: We have large contracts still being delivered against, but the pace is determined by end users. We are moving beyond initial customers as use cases are perfected and relevant to more customers. AI inference is growing rapidly, and our hybridized platform with AI services is seeing momentum.

Q: Regarding Nokia, what are kind of the next steps here, especially announceable steps in this relationship, and when you ultimately look forward to be contributing to backlog and eventually revenues?

A: We are building a joint solution, an AI platform focused on inference needs into their customers and ours. We will demonstrate and launch at Gitex Asia in the near future. There will be joint go - to - market co - selling into their customers, system integrators, cloud service providers, and enterprises. Revenues from the AI services platform, including Nokia, are towards the end of this year.

Q: Talk a little bit more about that broader opportunity set and if there's similar infrastructure wins that could come up that are along the lines of this deal with Nokia?

A: We do have similar opportunities in other continents. Asia has momentum, Africa has initial traction, and the US has massive infrastructure happening with a need for hybridized AI to serve business outcomes. The average model size is shrinking, and our graph streaming architecture in combination with GPUs is a perfect fit for these outcomes across various use cases.