The deal, still in early discussions, would give current and former employees an opportunity to cash out shares worth several billion dollars ahead of any initial public offering.

If finalised, the proposed valuation for OpenAI would be a sharp jump from the company’s existing $300 billion figure, stressing its rapid revenue growth and also the escalating race among global tech giants to attract and retain AI talent.

The surge has been driven largely by ChatGPT, OpenAI’s flagship product, which now has around 700 million weekly active users, up from 400 million in February and four times higher than last year. The company’s paying enterprise clients have grown to 5 million, compared to 3 million just two months ago.

Financially, OpenAI’s annual recurring revenue has climbed to $13 billion, from $10 billion in June, with expectations to hit $20 billion by year-end. The company doubled its revenue in the first seven months of the year and is expanding its product reach even as it navigates corporate changes.

The share sale discussions follow OpenAI’s recent $8.3 billion funding injection from investors including Dragoneer, Andreessen Horowitz, Sequoia, and Fidelity. 

This forms part of a $40 billion fundraising round led by Japan’s SoftBank Group, which has until the end of the year to complete its $22.5 billion commitment. The remainder has already been taken up at a $300 billion valuation, the source said.

The development is similar to private share sales by other high-growth technology firms such as ByteDance, Databricks, and Ramp, which have used such transactions to update market valuations and reward long-standing employees. 

Existing OpenAI investors, including Thrive Capital, are said to be in talks to participate in the sale. Thrive declined to comment.

Competition for AI talent is at an all-time high. Meta, for example, is investing billions in Scale AI in a bid to lure its 28-year-old CEO, Alexandr Wang, to lead its new superintelligence division. Across the industry, lucrative compensation offers are becoming the norm for top engineers and researchers.

OpenAI, meanwhile, is preparing for some structural changes. Plans are underway to move away from its capped-profit model, potentially clearing the way for a public listing. Chief Financial Officer Sarah Friar said in May that an IPO would happen only when both “the company and markets were ready.”

The company has also taken steps to reconnect with the open-source community, releasing open-weight models for the first time since 2019, a strategic move to counter competitors such as Anthropic.

A $500 billion valuation would place OpenAI among the most valuable privately held technology companies in history, rivalled only by firms like ByteDance and SpaceX. 

For the AI sector, it would be another sign that the biggest players are not limited to building powerful systems but are also securing the financial firepower to dominate the global market for years to come.

The announcement came directly from Omar Sanseviero, a developer relations engineer at Google DeepMind, who confirmed the figures on X over the weekend. 

In addition to the downloads, more than 70,000 custom variations of Gemma have already been built on Hugging Face, a widely used platform among developers.

Gemma, which launched in February 2024, is part of Google’s initiative to compete with other open model providers, especially Meta’s Llama series. 

The latest versions of Gemma can work with both images and text, a capability known as being “multimodal.” They also support more than 100 languages and include versions optimised for tasks like drug research.

However, when placed side-by-side with Meta’s Llama, Gemma’s traction seems to be on a moderate level. Llama passed 1.2 billion downloads in April 2025, nearly ten times what Gemma has achieved. In a space where momentum matters, such a gap is hard to ignore.

But volume alone doesn’t determine usefulness. What’s more important is what developers can actually do with these models. And that’s where the trouble starts. 

Both Gemma and Llama have been flagged by developers for having restrictive and unclear licensing. These are not the standard open licenses many in the community are used to. Some have argued that the vague terms make it risky to deploy the models in commercial settings.

That has led to hesitation, particularly among startups and businesses that can’t afford legal ambiguity. I’ve spoken with teams who say they’ve skipped over both Gemma and Llama entirely for this reason. The concern isn’t about capability, but clarity.

So while Google celebrates 150 million downloads, there are issues about what those downloads actually represent. Are developers using the models in real products? Or are they just experimenting, unsure whether they can actually go to market with them?