AI Trends

Nvidia Reportedly Plans to Spend Over $500 Million to Acquire Inception

ITHome · Artificial IntelligenceUpdated Nov 20 22:06

AI summary

The Information reports that Nvidia is in advanced talks to acquire AI startup Inception in a deal valued at more than $500 million, potentially with additional performance-based incentives. Inception develops Mercury, a diffusion-based large language model that it claims has a time to first token of under 300 milliseconds, delivers five to seven times the throughput of traditional models, and can reduce per-task costs by as much as 70%. Neither party has commented publicly, and the transaction price, closing date, and whether Mercury will continue to operate independently have yet to be determined.

Why read this

AI score 56

If the acquisition goes through, Nvidia will gain a team specializing in diffusion-based large language models and the Mercury technology, opening a new avenue for its open-source models and physical AI strategy.

Topics

Tags

Body · AI translation

Nvidia Reportedly to Spend More Than $500 Million to Acquire Inception, Maker of “AI Faster Than ChatGPT”

November 20, 2026, 14:15 — ITHome — Guyuan

Thanks to ITHome readers Yugong Qima and Buyao Ya for submitting the tip!

ITHome reported on November 20 that tech publication The Information published an article today (November 20) saying Nvidia plans to acquire AI startup Inception. The two parties are now in advanced negotiations, with the deal valuing the company at more than $500 million (ITHome note: approximately RMB 3.346 billion at current exchange rates), potentially with additional performance-based payouts.

Amir Efrati is the executive editor of The Information.

According to publicly available information reviewed by ITHome, Inception was founded in 2024 and is headquartered in Palo Alto, California. It was co-founded by Stanford University professor Stefano Ermon, UCLA professor Aditya Grover, and Cornell University professor Volodymyr Kuleshov.

The company primarily develops “diffusion large language models” (dLLMs). In February 2025, it released Mercury, which it describes as the world’s first commercial-scale diffusion large language model. Its products are available to developers through its own API, Azure AI Foundry, OpenRouter, Baseten, and other channels.

The Mercury model’s biggest highlight is parallel generation. Unlike the “autoregressive” approach used by OpenAI’s GPT series and Anthropic’s Claude series, the model employs a diffusion architecture: it begins with a rough passage resembling “noise,” then generates multiple tokens in parallel through repeated rounds of denoising and refinement.

In terms of cost-effectiveness, Inception says Mercury can achieve first-token latency of under 300 milliseconds, throughput five to seven times that of conventional models, and reductions of up to 70% in the cost of each task. The publication described it as “AI faster than ChatGPT.”

According to the report, if the deal is completed, Inception’s three co-founders are expected to join Nvidia. Its team may also contribute to Nvidia’s Nemotron open-source model ecosystem and extend diffusion technology into robotics and other “physical AI” fields.

Neither Nvidia nor Inception has publicly commented. As negotiations have yet to be finalized, the transaction price, closing date, and whether products such as Mercury will continue to operate independently all remain uncertain.

Source

ITHome · Artificial Intelligence