Urgent.News

What's breaking now, across thousands of outlets.

AI

OpenAI Q3 revenue soars nearly 50% to US$10 billion

It is up nearly 50% from the previous quarter and almost three times a year earlier.

OpenAI reported a staggering US$10 billion (S$12.8 billion) in revenue for the third quarter ending in September, marking a remarkable 50% increase from the previous quarter and nearly three times the revenue from the same period a year ago, according to a person familiar with the company's financials. This surge in revenue comes after OpenAI has recently gained traction with new AI models, competitive pricing, and a stronger emphasis on AI agents to automate tasks for customers.

The company's recent financial results, shared with investors during a call on November 10, indicate a resurgence in its business performance as it tries to catch up with its long-standing rival, Anthropic. OpenAI's revenue growth outpaced Anthropic's, which reported a slower third-quarter revenue growth of around 47%, from the second quarter.

Both companies have been navigating the competitive landscape in the AI sector, with OpenAI nearing a significant funding round at a valuation of US$1.25 trillion, while Anthropic is aiming for a US$1.4 trillion valuation for its public listing later in November.

Written by urgent.news from Straits Times Business's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

This story

This is one outlet's version. Read the fullest account.

Read the original at straitstimes.com →

More in AI

Why vLLM and SGLang might not be the best option to run AI agents

vLLM and SGLang are great inference engines. If you serve chat or batch jobs, use them. But AI agents are a different workload, and I wanted to see how well these engines handle it. So I measured it.

  • vLLM and SGLang may not be optimal for AI agents due to cache management issues.
  • Rust session engine reduces history recomputation to 8-16% of total tokens.
  • Rust session engine produces first token about 5 times faster than SGLang and vLLM.

The Silent Costs of AI APIs Nobody Warns You About

I remember the exact moment my AI API bill made me choke on my coffee. It wasn't a massive spike from heavy traffic. It was a $340 charge for a weekend where I barely touched the codebase.

  • Tokenization process inflates token count, leading to higher costs.
  • Conversation history and rate limits create unexpected charges.
  • Output tokens are more expensive than input tokens, disproportionately impacting bills.

More from Wednesday 11 November →