Nvidia is in advanced talks to acquire Inception, a startup that builds diffusion-based language models, for more than $500 million plus earnouts, The Information reported on 19 November, citing sources.
The talks began before Nvidia’s chief executive resigned, the outlet said. Neither company has announced a deal.
Important: not investment advice
This article is for information and education only. It is not investment, financial, legal or tax advice, and it is not a recommendation or an offer to buy or sell any security, ETF, option or other financial product. FourWeekMBA is not a broker-dealer or a registered investment adviser. IPO shares, leveraged and inverse ETFs, and options carry significant risk, including the loss of your entire investment; leveraged and inverse ETFs reset daily and can lose value quickly. Prices, availability, fees and eligibility change and can differ by country and broker, so check the prospectus and your broker’s terms. Figures marked “our arithmetic” are our own calculations from the cited filings and may contain errors. Do your own research and consider speaking to a licensed financial professional before making any investment decision.
Business Pill · Vertical Integration
An evergreen Business Engineer explainer on vertical integration. It teaches the general idea only and does not discuss Nvidia’s talks with Inception.
The key insight: The Information reports Nvidia is in advanced talks to buy diffusion-model startup Inception for more than $500 million plus earnouts, over 10 times the $50 million seed round its own NVentures joined.
What Inception Builds
Inception is a two-year-old company led by Stanford professor Stefano Ermon. Its co-founders are Ermon, Aditya Grover and Volodymyr Kuleshov.
Most chatbots write text one piece at a time. Inception’s diffusion models instead “modif[y] the overall structure of a response incrementally” until it matches the result, TechCrunch reported.
Its first model family is called Mercury. TechCrunch said Mercury had been integrated into developer tools including ProxyAI, Buildglare and Kilo Code.
In February, Inception launched Mercury 2. The company said it reaches 1,009 tokens per second on Nvidia Blackwell GPUs, The Decoder reported.
Inception put end-to-end latency at 1.7 seconds, against 14.4 seconds for Gemini 3 Flash and 23.4 seconds for Claude Haiku 4.5 with reasoning on. Those are the company’s figures.


Nvidia Already Owns a Piece
Inception raised $50 million in seed funding in November 2025, led by Menlo Ventures. NVentures, Nvidia’s venture arm, took part, alongside Microsoft’s M12, Mayfield, Innovation Endeavors, Snowflake Ventures and Databricks Investment.
Andrew Ng and Andrej Karpathy added angel money. We found no later funding round or valuation in public sources.
A price above $500 million would be more than 10 times the $50 million seed round (our arithmetic). The earnouts would come on top.
Why Diffusion Matters to Nvidia
Speed is the pitch. A model that answers in under two seconds is cheaper to serve per answer, which matters to anyone selling inference.
Mercury 2’s headline numbers were set on Nvidia Blackwell chips. Owning the model team would give Nvidia its own showcase for fast inference on its hardware.
The talks come in a hard week for Nvidia. California’s attorney general subpoenaed the company on Thursday, and the stock closed below its listed 52-week range.
The Structural Read
As we read it, Nvidia would be buying a team whose selling point is fast, cheap inference on Nvidia chips.
As we read the funding record, Nvidia would be acquiring a company it already invested in, which is a common route for corporate venture arms.
TechCrunch on Inception’s diffusion models, 6 Nov 2025
“modif[y] the overall structure of a response incrementally”
Three Implications
FOR NVIDIA A model team built around inference speed on Blackwell, if the talks close.
FOR INCEPTION’S INVESTORS A reported price more than 10 times the $50 million seed round, before earnouts.
FOR AI LABS A chipmaker owning a model maker adds to the overlap between suppliers and customers.
What Is Not Established
We could not read The Information’s article, which is paywalled. We do not know the earnout terms, the deal structure or whether terms are signed.
Nvidia and Inception had not commented in the sources we read. Talks can still fail.
The Bottom Line
Nvidia is reported to be in advanced talks to buy a fast-inference model maker its venture arm already backed, at more than 10 times its $50 million seed round.
Important: not investment advice
This article is for information and education only. It is not investment, financial, legal or tax advice, and it is not a recommendation or an offer to buy or sell any security, ETF, option or other financial product. FourWeekMBA is not a broker-dealer or a registered investment adviser. IPO shares, leveraged and inverse ETFs, and options carry significant risk, including the loss of your entire investment; leveraged and inverse ETFs reset daily and can lose value quickly. Prices, availability, fees and eligibility change and can differ by country and broker, so check the prospectus and your broker’s terms. Figures marked “our arithmetic” are our own calculations from the cited filings and may contain errors. Do your own research and consider speaking to a licensed financial professional before making any investment decision.
97,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. The report of the acquisition talks, the price and the earnouts is The Information’s; we read only Techmeme’s summary, as the article is paywalled. Funding details are from TechCrunch and Inception’s November 2025 release. Mercury 2 speed, latency and comparisons are Inception’s claims as reported by The Decoder. The 10-times multiple is our arithmetic.
Sources: The Information via Techmeme, “Nvidia is in advanced talks to acquire Inception,” 19 Nov 2026 · TechCrunch, “Inception raises $50 million to build diffusion models for code and text,” 6 Nov 2025 · The Decoder, “Inception launches Mercury 2,” 24 Feb 2026









