OpenAI to buy cloud startup Ona to support artificial intelligence agents

OpenAI has struck a deal to acquire Ona, a startup specializing in cloud-based infrastructure designed to support artificial intelligence agents, as the company looks to strengthen its offerings for enterprise customers and expand the practical capabilities of its AI tools.

The acquisition, which is yet to be finalized, will see Ona’s team join OpenAI’s Codex initiative, according to an announcement expected from the ChatGPT creator on Thursday.

OpenAI also revealed that more than 5 million users are now using its Codex AI coding tools every week, highlighting the growing adoption of its developer-focused products.

Ona develops secure and persistent computing environments that allow AI agents to access the tools, systems and contextual information needed to perform tasks and improve over time.

OpenAI said the technology will help create more capable and reliable AI agents for business applications.

The move comes as OpenAI intensifies its rivalry with Anthropic, with both companies racing to build advanced AI systems that can automate complex workplace tasks and attract corporate customers. The competition has also extended to capital markets, with both startups reportedly filing confidential paperwork for potential initial public offerings and considering stock market debuts as early as this fall.

Financial details of the acquisition were not disclosed.

The company behind ChatGPT confidentially submitted paperwork for an initial public offering (IPO) earlier this week, following a similar filing by rival Anthropic last week. OpenAI CEO Sam Altman has previously highlighted the challenge of rising operational expenses as one of the company's biggest concerns.

The potential price cuts may be aimed at staying ahead of Anthropic, which is also believed to be evaluating similar measures, the report said. The two firms are locked in a fierce race for dominance in the enterprise AI segment. Anthropic's Claude platform has gained traction among corporate customers, with the company targeting second-quarter revenue of USD 10.9 billion, up sharply from USD 4.8 billion in the previous quarter.

"The company is weighing significant cuts to what it charges for tokens, the unit of measurement artificial-intelligence firms use to bill for their products," the WSJ report said.

Competition is also intensifying in the AI-assisted coding space, where Anthropic's Claude Code and OpenAI's Codex are vying to attract developers and software engineering teams.

Anthropic recently overtook OpenAI in valuation after securing USD 65 billion in its Series H funding round in May, valuing the company at USD 965 billion. By comparison, OpenAI was valued at USD 852 billion during its last fundraising round in March.

The rivalry has escalated as businesses worldwide increasingly adopt agentic AI tools to improve efficiency and automate workflows. However, any aggressive pricing strategy could place additional pressure on profitability, as both companies continue to spend heavily on computing infrastructure required to train and run advanced AI models.

Meanwhile, another major AI-focused player, SpaceX, is set to debut on the Nasdaq on Friday. The company, owned by Elon Musk, has priced its IPO at USD 135 per share, seeking a valuation of USD 1.75 trillion. A strong market debut could further boost Musk's wealth and strengthen his position among the world's richest individuals.

Across the industry, AI giants are investing billions of dollars to expand capacity in anticipation of surging demand. Leading cloud providers and hyperscalers, including Amazon AWS and Alphabet-owned Google, are also ramping up investments and fundraising efforts to build larger compute infrastructure for the next wave of AI growth.