i wrote it 2 min after seeing the post on X, and linked to the source. wild that you guys let other outlets do the same, but mine always gets "post the source"
Edit: before you say I just regurgitated the release, not true. Here's the exact color I added:
### Context and market implications
Microsoft’s claims arrive at a time when the industry is grappling with the escalating cost of running large foundation models. OpenAI’s GPT‑5 series and Anthropic’s Claude 5 have set new performance baselines but require high‑end GPUs and generate sizable token bills. By delivering comparable performance in Excel on A100 hardware, Microsoft demonstrates a potential path to more sustainable AI in everyday software.
However, the lack of third‑party benchmarks leaves open the question of how these MAI models fare against the latest open‑source alternatives such as Llama‑3‑70B or MosaicML’s MPT‑30B, which also claim strong cost‑performance profiles. Microsoft does not disclose the exact token‑per‑query cost reduction, making it hard to quantify the economic impact for enterprise customers.
If the hill‑climbing framework proves scalable across the broader Microsoft 365 suite, it could shift the balance of power from external AI providers toward in‑house, product‑tailored models. That would give Microsoft greater control over data, pricing, and roadmap while potentially reducing the marketplace share of OpenAI and Anthropic in the enterprise segment.
### What the numbers say (and don’t say)
* *Code‑accept rate*: 10% higher than GPT‑5.4 Mini and Claude Haiku 4.5 – internal metric; no public benchmark.
* *Token usage*: 10% lower median tokens – again internal telemetry.
* *Performance parity*: "On par with GPT‑5.6" in Excel – no disclosed evaluation criteria.
The blog post does not disclose the total cost savings in dollar terms, nor does it reveal the latency impact of running the model on A100 versus H100. Those gaps matter for enterprise procurement decisions, which often hinge on total cost of ownership.
### Why it matters
Microsoft is turning its massive product ecosystem into a training ground for bespoke, cost‑efficient AI models. If the hill‑climbing approach scales, it could redefine how enterprise software vendors build and monetize AI, reducing reliance on pricey external foundation models and tightening the feedback loop between user behavior and model improvement.
Yup and not the same news. The point of this article was to show that the neoclouds are not going away, complete with timelines and context on how Meta is beholden to the neoclouds. Totally different angle.
can you be more specific? because i specifically make sure we don't do those types of parallel semantics in our agent and articles. and i dont see any parallel constructions in this write up?
edit: nevermind, i see it. cleaned it up. thanks for the feedback!
No agency. Just a one man shop. What kind of vibecoded website is this? It's a Replit built site along wtih my 20 years of experience building websites from scratch. It was launched this month, so it's a bit rough around the edges, but we're off to a great start.
no model. that was me. i should have confirmed they were the same before posting. when i assumed they were the same, the real Wei's credentials made the language much more exciting. apologies. i'll do better next time.
Sure you can. You just need enough land and money to start basic things like a post office, city hall, courthouse, roads, and a way to get power to the whole thing.