Up Close
Microsoft folds Copilot into one client as rivals move into Office and cheaper fast models
Microsoft said it is starting to roll out one Copilot app for personal, work and school accounts, merging its consumer Copilot and Microsoft 365 Copilot clients and retiring several consumer-only features.
That looks like more than a product tidy-up. It likely means Microsoft now needs Copilot to behave as one surface across private and paid use, because the boundary between those contexts is becoming less useful than the continuity between them. AWS released Amazon Quick extensions inside Word, Excel, PowerPoint and Outlook for existing customers without extra licensing. That is likely to weaken the old assumption that Microsoft can reserve the Microsoft 365 desktop as Copilot's natural home.
Google DeepMind introduced Gemini 3.7 Flash and said it launches at roughly half the price per million tokens of Gemini 3.6 Flash, while OpenAI introduced an Ultrafast API tier for GPT-5.6 Sol in preview for time-sensitive applications. Taken together, this suggests that model competition is pressing on price and latency at the same time, not only on raw capability. For Microsoft, that likely shifts more of the burden of differentiation from the model itself to packaging, account control and where the assistant appears by default.
Evidence
- 2026-08-13 Microsoft announced it is beginning to roll out a unified Copilot app that combines its consumer Copilot and Microsoft 365 Copilot into a single client supporting personal, work and school accounts, while retiring several consumer‑only features. thurrott.com
- 2026-08-13 AWS released Amazon Quick extensions that embed its agentic AI assistant directly into Microsoft Word, Excel, PowerPoint and Outlook for existing Plus, Professional and Enterprise customers without additional licensing. aws.amazon.com
- 2026-08-13 Google DeepMind unveiled Gemini 3.7 Flash, a new workhorse model optimized for coding and agentic workflows that improves benchmark performance and web/app generation quality while launching at roughly half the price per million tokens of Gemini 3.6 Flash. blog.google
- 2026-08-13 OpenAI introduced Ultrafast, a new OpenAI API service tier that runs GPT‑5.6 Sol up to 14× faster than standard processing and is launching in preview for time‑sensitive applications such as incident response and real‑time customer support. openai.com