OpenAI Raises $122B and Launches Real-Time Voice AI Models
OpenAI closed a $122B funding round anchored by Amazon, Nvidia, SoftBank, and Microsoft, and launched three new real-time audio AI models for enterprise.
By TechPopDaily Admin
Updated May 18, 2026
OpenAI Is Now Worth More Than Most Countries GDP
OpenAI closed a $122 billion raise anchored by Amazon, Nvidia, SoftBank, and Microsoft, valuing the company at $852 billion. To put that in perspective: OpenAI is now worth more than the entire GDP of Switzerland. The ChatGPT maker — founded in 2015 as a nonprofit research lab — has become one of the most valuable companies in the world in less than three years of consumer product launches.
The Strategic Logic Behind the $122B Mega-Round
Amazon's participation ensures deep integration with AWS, making OpenAI models available to millions of enterprise customers through cloud APIs they already use. Nvidia's investment is a bet on a customer that consumes more H100 and Blackwell GPUs than almost any other organization on earth. Microsoft's ongoing participation reinforces the Azure partnership powering ChatGPT and the Copilot suite. The $852 billion valuation means OpenAI is now more valuable than Johnson and Johnson, HSBC, and Alibaba combined. A trillion-dollar valuation and IPO appears inevitable within 18 months.
GPT-Realtime-2: Conversational AI That Gets Things Done
Alongside the funding, OpenAI unveiled three new real-time audio models for enterprise deployment. GPT-Realtime-2 is designed for conversational task execution — a voice AI that does not just answer questions but completes workflows in real time. Enterprise use cases include AI-powered call center agents handling customer support end-to-end without human escalation, voice-driven internal helpdesks that query databases and create tickets in a single conversation, and real-time sales coaching tools that listen to calls and surface objection-handling suggestions as conversations unfold. The key technical advance is dramatically reduced latency — under 200 milliseconds round-trip — with improved handling of interruptions, accents, and domain-specific vocabulary.
GPT-Realtime-Translate: 70 Languages, Real Time
The second model handles multilingual translation across more than 70 languages in real time. This is not basic translation — it is context-aware, culturally calibrated live translation that preserves tone, formality, and industry terminology. Global companies can deploy a single AI model speaking 70 languages fluently, without maintaining separate localized customer support teams for every market. For the 1.8 billion English learners worldwide, real-time translation reduces the language premium currently disadvantaging non-English speakers in global commerce.
GPT-Realtime-Whisper: Live Transcription at Enterprise Scale
The third model handles live transcription and captioning with enterprise-grade accuracy. Built on OpenAI's Whisper technology but redesigned for real-time output, it enables meeting transcription, accessibility captioning, compliance recording, and searchable voice archives at scale. For regulated industries — finance, healthcare, legal services — where documentation of conversations is mandatory, GPT-Realtime-Whisper offers a transformative alternative to expensive manual transcription. A hedge fund compliance team could automatically generate searchable, timestamped records of every analyst call.
OpenAI's Government and Infrastructure Ambitions
The funding comes alongside OpenAI making its most advanced AI models available to all vetted levels of the US government — positioning itself as the de facto AI standard for federal agencies, a market worth hundreds of billions in long-term contracts. With $122 billion in fresh capital, three new enterprise voice products, and government access, OpenAI in 2026 is not a startup that made a chatbot. It is a full-stack AI infrastructure company becoming as fundamental to business operations as the internet itself.
Palo Alto warns of a 3-5 month window before AI-driven cyberattacks become the new norm. 87% of security teams already see more AI threats than ever before.