
Image via TechCrunch
Wednesday, April 1, 2026
OpenAI raises $122B at $852B valuation
OpenAI raised $122 billion at an $852 billion valuation to expand infrastructure, while its Sora video model brought in $2.1 million in revenue against $1 million in daily costs. Google launched Veo 3.1 Lite at half the price of its standard model.

Image via TechCrunch
Top Stories
OpenAI
OpenAI raised $122 billion at an $852 billion valuation to expand compute infrastructure and enterprise AI capabilities, reflecting unprecedented investor confidence in the company's growth trajectory and market position.
The Wall Street Journal
Sora generated just $2.1 million in revenue over six months while costing $1-15 million daily to operate, exposing the severe economic challenge of commercializing expensive AI video generation models at scale.
apoorv03.com
The AI value chain remains severely inverted with NVIDIA and semiconductors capturing 70% of revenue and 79% of gross profit in a $435B ecosystem, while applications (dominated by OpenAI/Anthropic) earn just $60B at lower margins. Despite 5x growth in two years, custom silicon efforts from hyperscalers may take over a decade to shift economics toward the application layer.
Google Blog
Google released Veo 3.1 Lite, a video generation model offering the same speed as Veo 3.1 Fast at less than half the cost, targeting developers building high-volume video applications. The launch expands Google's video model family and includes upcoming price reductions across the Veo lineup.
Meta Newsroom
Meta is releasing prescription-first Ray-Ban Meta smart glasses starting at $499 with enhanced fit customization, while adding multimodal AI features including hands-free nutrition tracking, on-device WhatsApp summaries, and expanded navigation capabilities. The move signals Meta's push to make AI-powered wearables more accessible to the broader prescription eyewear market.
Keep Reading
Industry Voices
Bingchen Zhao
Shares cutting-edge research on visual recognition and efficient model training from his work at Edinburgh and projects like OTTER for instruction-tuned vision models.
Wannan Yang
Posts insights on large-scale recommender systems and ML infrastructure from experience building personalization algorithms at major tech companies.
Enjoyed this issue?
Get daily AI intel delivered to your inbox. No fluff, just the stories that matter.