0
Applied AI·July 23, 2026·1 min read

Microsoft launches new in-house AI models it says cut costs up to 89% versus OpenAI

Share

If Microsoft can show production workloads running 50–89% cheaper on MAI-Image-2.5-Pro and MAI-Voice-2-Flash than on OpenAI, the internal-vs-partner model debate just became a hard cost line item. Enterprise AI teams should re-open their TCO spreadsheets this week and model scenarios where foundation vendors are also your direct price competitors.

Applied AI

AMD is working with Cerebras to combine AMD Helios and Cerebras wafer-scale chips for an AI inference solution to make workloads faster; CBRS closed up 4.86%

AMD teaming up with Cerebras on a combined Helios + wafer-scale inference rack underscores how heterogeneous AI compute is becoming at the system level, not just the chip level. Infra buyers should start architecting for mixed-vendor, mixed-accelerator clusters rather than assuming a single GPU family will dominate their next buildout.