OpenAI says its new Jalapeño AI chip could significantly reduce the cost of running artificial intelligence by delivering substantially higher efficiency and faster inference, but Jim Cramer remains skeptical that the chip threatens Nvidia Corp’s (NASDAQ:NVDA) dominance.

OpenAI’s Jalapeño Chip Targets Lower AI Costs

On Tuesday, OpenAI said Jalapeño, its first custom inference chip developed with Broadcom Inc. (NASDAQ:AVGO), is designed to deliver both high throughput and low latency—two areas that are typical trade-offs.

Since announcing Jalapeño, our first custom inference chip, we’ve been testing it and the system around it.

The results show a major advance: more intelligence from every watt and faster responses, delivering both higher throughput and lower latency in one architecture without… pic.twitter.com/vj7VOrA8pP

— OpenAI (@OpenAI) August 25, 2026

The company plans to begin deploying Jalapeño within its infrastructure by the end of 2026, with production expected to ramp further in 2027.

OpenAI is already working on second- and third-generation versions, OpenAI hardware Vice President Richard Ho told reporters, according to The Verge.

However, OpenAI does not view Jalapeño as an outright Nvidia replacement. The company said it will continue deploying accelerators from Nvidia and other partners for training and inference.

OpenAI CEO Sam Altman took to X  and said, "We made a chip and it is fast."

we made a chip and it is fast

— Sam Altman (@sama) August 25, 2026

Jalapeño Beats Nvidia on AI Inference Benchmarks: OpenAI

OpenAI’s own benchmark results showed Jalapeño delivering 1.5x to 1.9x more AI work per watt and 1.7x to 3.6x lower end-to-end latency than comparison systems based on Nvidia’s GB200 and GB300 chips across GPT-OSS 120B, DeepSeek R1 and Kimi K2.5 1T.

The chip is designed specifically for inference, or the process of running trained AI models to generate responses and complete tasks.

OpenAI’s InferenceX benchmarks show Jalapeño outperforming Nvidia’s GB200 and GB300 across all three AI models, with higher performance per watt and significantly lower latency and token-to-token time.

AI Model	NVIDIA Comparison	Peak Performance Per Watt	End-to-End Latency	Minimum TBT
GPT-OSS 120B	GB200	1.9x higher	1.7x lower	2.7x lower
DeepSeek R1 670B	GB300	1.7x higher	3.6x lower	4.1x lower
Kimi K2.5 1T	GB300	1.5x higher	3.4x lower	3.8x lower

Nvidia did not immediately respond to Benzinga’s request for comment.

OpenAI’s Jalapeño Chip Could Lower AI Costs

In a conversation with Bloomberg TV, Ho said performance per watt could be 1.8x to 4x better than existing chips, potentially allowing the company to lower the cost of serving AI models as Jalapeño scales into production.

"Ultimately, when this rolls out into production, customer token prices should come down," Ho said.

OpenAI says its new Jalapeño chip could dramatically cut the cost of AI.

The company's VP of hardware Richard Ho says performance per watt is 1.8x to 4x better than existing chips – potentially translating into cheaper tokens for customers as it rolls into production… pic.twitter.com/0kMqHNGu95

— Bloomberg TV (@BloombergTV) August 25, 2026

Jim Cramer Dismisses Nvidia Challenger Claims

Cramer, however, isn’t convinced another custom AI chip will seriously challenge Nvidia.

"Every day I read about some chip that is superior to Nvidia. And every year I see no real competitors," Cramer said in a post on X.

Every day i read about some chip that is superior to Nvidia. And every year i see no real competitors

— Jim Cramer (@jimcramer) August 25, 2026

Price Action: Nvidia shares closed at $213.05, up 2.19% on Tuesday. The stock rose another 0.31% to $213.70 in after-hours trading, according to Benzinga Pro.

According to Benzinga Edge Rankings, Nvidia ranks in the 99th percentile for growth and maintains positive price trend ratings across the short, medium and long term.

Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.

Photo Courtesy: katz / Shutterstock.com