CNBC
02 Jun 2026, 18:34 UTC · 2h ago
Microsoft unveils new AI models to lessen reliance on OpenAI and lower costs for developers
Search tags
CNBC
02 Jun 2026, 18:34 UTC · 2h ago
Search tags
Story key points
4 claims · impact-rated
Microsoft has launched its own proprietary AI models, MAI-Code-1-Flash and MAI-Thinking-1, to compete directly with frontier models from partners like OpenAI and Anthropic. — Reducing reliance on third-party models allows Microsoft to capture more margin and control its AI stack, though it creates a competitive conflict with its major investees.
+0.60Microsoft claims its refined models outperformed OpenAI's GPT 4o (referenced as GPT 5-5 in text) for McKinsey with 10 times better cost efficiency. — Direct evidence of superior cost-performance suggests a potential shift in enterprise adoption toward Microsoft's native models over OpenAI's.
+0.50Anthropic has confidentially filed for an IPO, and OpenAI is pursuing a potential offering this year. — Public listings for these AI giants would provide a benchmark for AI valuations and potentially trigger significant capital inflows into the sector.
+0.40Continue reading
6 related stories
Microsoft is introducing small Aion models capable of running locally on Windows PCs. — Edge AI deployment drives demand for AI-capable hardware and increases the utility of the Windows ecosystem.
+0.30Ticker attribution
Model heads
Launching new proprietary AI models to reduce reliance on third parties and improve cost efficiency for developers.
Mentioned as a competitor with its Gemini 3.5 Flash model, but no new positive or negative developments were reported.
Early access
News Impact Screener scores every headline against the stocks it moves — before the chart reacts. Join the early-access list and get alerted the moment a story hits your tickers.
Impact vectors
4 dimensions · 9 clusters
Market reaction
10 bid · 10 offered
Bloomberg Technology
1h ago
