OpenAI told investors that it reached roughly $50 billion in annualized revenue at the end of September, down from the $68 billion figure widely reported late last month.

The $68 billion figure included gross revenue from OpenAI’s partners, which helps investors make a more direct comparison with its chief rival, Anthropic.

In addition, OpenAI touted 77% total run-rate growth in its third quarter and 107% run-rate growth in its enterprise business during the same period.

After the market learned more details about OpenAI’s revenue, AI stocks, including Nvidia, Oracle and CoreWeave, sank. Nvidia shares fell 3%, Oracle shares fell nearly 6%, and CoreWeave shares slipped nearly 8% yesterday. Additionally, Advanced Micro Devices fell 4%, Broadcom fell 4%, Intel fell 5%, and Super Micro Computer fell nearly 5%.

OpenAI is under pressure to justify its $852 billion valuation to investors as it gears up for its IPO. The company closed a historic $122 billion funding round in March.

Anthropic is also preparing for a major IPO. The company has not officially disclosed when it plans to debut, but it’s been meeting with prospective investors and is reportedly seeking a $2 trillion valuation.

In August, Anthropic told investors that its annualized revenue run rate reached $65 billion as of the end of July.

Source: Youtube

Source: Youtube

Source: Youtube

Source: Youtube

Source: Youtube

Reflection AI, a New York AI startup that presents itself as the U.S.’s answer to open-source Chinese AI, released its first model, Beam, claiming it outperforms Western open models.

This company targets a large market of businesses, governments, and other entities building sovereign AI systems that don’t want to use Chinese models.

Founded in 2024 by former DeepMind researchers, Reflection AI said that the model is particularly strong at coding and agentic tasks. It also needs three to four times less computing power to reason, making it faster and cheaper to run.

The cofounder and CEO, Misha Laskin, said the model’s capabilities are comparable to China’s Z.ai’s recent model, GLM-5.2, and Alibaba’s most advanced model, Qwen 3.8-Max — which trails the top closed American models from OpenAI and Anthropic.

Several large Western companies have recently adopted Chinese models to reduce costs for certain repetitive tasks, such as customer service and writing routine code. Beijing is also promoting open source as a diplomatic tool to help other countries in their tech development.

However, because open models can be downloaded and run with fewer safeguards, their rise has also raised security concerns: US and UK government evaluators found that recent Chinese models could help hackers exploit code vulnerabilities.

Source: Youtube

Source: Youtube

Source: Youtube