Private Markets, Angel and Secondary Investing, Sacra

Sacra

Private Markets, Angel and Secondary Investing

Bringing transparency to private markets so investors, founders, and employees can build on conviction. Sacra delivers deeply researched insights on the private companies shaping tomorrow's markets. By surfacing the data, context, and operator perspective that are missing from private markets today, we help secondary buyers and sellers price equity accurately and unlock liquidity for the people building these businesses.

Private Capital Markets

<ul><li>Sacra estimates that OpenRouter hit $5M in annualized revenue in May 2025, on $100M in GMV, up from $1M in annualized revenue at the end of 2024.</li><li>OpenRouter customers buy tokens for inference that get spent across various providers like OpenAI and Anthropic, with tokens marked up roughly 5%—with the value prop for developers being faster setup, a simpler integration experience, and the ability to abstract away the complexity of model selection, with one API, one dashboard for usage & billing, and cost savings from automatically down-shifting to cheaper models when possible.</li><li>The explosion of AI models across 60+ providers and 400+ model variants is making OpenRouter core infrastructure for connecting AI apps to models.</li></ul>

Sacra estimates that OpenRouter hit $5M in annualized revenue in May 2025, on $100M in GMV, up from $1M in annualized revenue at the end of 2024.OpenRouter customers buy tokens for inference that get spent across various providers like OpenAI and Anthropic, with tokens marked up roughly 5%—with the value prop for developers being faster setup, a simpler integration experience, and the ability to abstract away the complexity of model selection, with one API, one dashboard for usage & billing, and cost savings from automatically down-shifting to cheaper models when possible.The explosion of AI models across 60+ providers and 400+ model variants is making OpenRouter core infrastructure for connecting AI apps to models.

<ul><li>Lightmatter reached a $4.4 billion valuation in October 2024 following a $400 million Series D round led by T. Rowe Price.</li><li>The company has raised a total of $850 million since its 2017 founding through multiple funding rounds.</li><li>Lightmatter's go-to-market strategy focuses on securing large deals with major cloud and AI data center operators, rather than pursuing smaller end-users.</li></ul>

Lightmatter reached a $4.4 billion valuation in October 2024 following a $400 million Series D round led by T. Rowe Price.The company has raised a total of $850 million since its 2017 founding through multiple funding rounds.Lightmatter's go-to-market strategy focuses on securing large deals with major cloud and AI data center operators, rather than pursuing smaller end-users.

<ul><li>The MLOps landscape has evolved from targeting researchers who preferred simple Python APIs and quick setup to serving DevOps professionals who need enterprise-grade telemetry, monitoring and deployment tools that can handle high-volume production workloads.</li><li>LLM deployment differs fundamentally from traditional software deployment because models are massive files (often 10-20GB) that must be loaded across CPU and GPU memory, creating new technical challenges around memory management, version control, and continuous deployment that existing DevOps tools weren't built to handle.</li><li>Building on microservices and serverless, compound AI systems merge together multiple custom AI models, LLMs and diffusion models with databases and external APIs—spawning a new generation of tools like Outerport and Modal ($32M raised) that let developers automatically deploy and run multi-LLM systems across CPU/GPU architectures without worrying about infrastructure.</li></ul>

The MLOps landscape has evolved from targeting researchers who preferred simple Python APIs and quick setup to serving DevOps professionals who need enterprise-grade telemetry, monitoring and deployment tools that can handle high-volume production workloads.LLM deployment differs fundamentally from traditional software deployment because models are massive files (often 10-20GB) that must be loaded across CPU and GPU memory, creating new technical challenges around memory management, version control, and continuous deployment that existing DevOps tools weren't built to handle.Building on microservices and serverless, compound AI systems merge together multiple custom AI models, LLMs and diffusion models with databases and external APIs—spawning a new generation of tools like Outerport and Modal ($32M raised) that let developers automatically deploy and run multi-LLM systems across CPU/GPU architectures without worrying about infrastructure.

Towaki Takikawa, CEO And co-founder Of Outerport, On The Rise Of DevOps For LLMs

Covering IPOs, Healthcare and Technology Stocks

Andrei Zakharov

Equity Capital Markets

Cerebras Systems / Equity Capital Markets

<ul>
<li>Cerebras Systems, a fast-growing maker of wafer-scale AI chips, which power the world’s largest supercomputers, will reveal IPO terms soon.</li>
<li>Given the company’s hyper-growth trajectory, $1B+ deal with UAE’s G42 and unique architecture, I expect Cerebras Systems is likely to get a premium valuation.  </li>
<li>The AI company plans to raise up to $1B while selling stockholders will offer additional Class A shares in this IPO.</li>
</ul>


Cerebras Systems, a fast-growing maker of wafer-scale AI chips, which power the world’s largest supercomputers, will reveal IPO terms soon.
Given the company’s hyper-growth trajectory, $1B+ deal with UAE’s G42 and unique architecture, I expect Cerebras Systems is likely to get a premium valuation.  
The AI company plans to raise up to $1B while selling stockholders will offer additional Class A shares in this IPO.


Cerebras IPO Valuation Analysis: A Sky-High Valuation Will Not Last For Long

<ul>
<li>Cerebras Systems, a high-growth Ai chip maker and hardware company that builds AI supercomputers with scalable architecture, has filed for an IPO on the Nasdaq.</li>
<li>The AI company plans to raise up to $1B in upcoming IPO. Citigroup and Barclays are the lead bankers on the offering. </li>
<li>Cerebras Systems may become the biggest disruptor since NVIDIA. The company’s growth is exceptionally high for any AI chip maker.</li>
</ul>


Cerebras Systems, a high-growth Ai chip maker and hardware company that builds AI supercomputers with scalable architecture, has filed for an IPO on the Nasdaq.
The AI company plans to raise up to $1B in upcoming IPO. Citigroup and Barclays are the lead bankers on the offering. 
Cerebras Systems may become the biggest disruptor since NVIDIA. The company’s growth is exceptionally high for any AI chip maker.


Cerebras IPO Preview: Does The Size Matter? Top-Tier VC Investors and $1B+ Deal With UAE’s G42

United States

Cerebras Systems Inc. is an artificial intelligence infrastructure company. The Company designs and manufactures an AI compute platform comprised of proprietary systems and software that is delivered in standard racks for deployment in data centers up to supercomputer scale. Cerebras Systems serves customers in the United States.

Price Gauge
52-week Range	0.00 - 0.00
Market Cap (USDm)	0.00
P/E	0.00
Dividend (Yield)	0.00

Cerebras Systems (CBRS US)

OpenRouter At $100M GMV

TL;DR: As developers move from wrapping single LLMs to stitching together multiple models, OpenRouter has emerged as the universal LLM...

Sacra

Primer: Lightmatter - 12 Feb 2025

Hardware and software stack for photonic AI accelerators and data center interconnects

Sacra

Towaki Takikawa, CEO And co-founder Of Outerport, On The Rise Of DevOps For LLMs

What does Outerport do...

Sacra

Cerebras IPO Valuation Analysis: A Sky-High Valuation Will Not Last For Long

In my follow-up piece, I discuss Cerebras Systems valuation analysis, including peers, multiples and the company’s revenues forecast for 2024 and...

Andrei Zakharov

Cerebras IPO Preview: Does The Size Matter? Top-Tier VC Investors and $1B+ Deal With UAE’s G42

Cerebras Systems, a manufacturer of the largest AI chip ever built and hardware company that launched supercomputers in the United States, filed...

Andrei Zakharov

Price Gauge

52-week Range

Market Cap (USDm)

P/E

Dividend (Yield)