One of fourteen models named Amazon WorkSpaces first on the direct prompt; one named Windows 365. Amazon WorkSpaces was named by fourteen of the fourteen models and Windows 365 by eleven and Amazon WorkSpaces carries 61 labels and Windows 365 32, so the shares are not directly comparable.
Named in one category this edition.
Named in two categories this edition.
Share is the count of first choices across the direct, paraphrase, budget and scale prompts over all fourteen models, for a mid-market B2B company; rank is within the category; every quote names the model and the prompt it came from. Both figures come from the virtual desktops page.
Across every category in the October 2026 Edition, Amazon WorkSpaces and Windows 365 were named in the same answer eighty times, of the 181 answers naming Amazon WorkSpaces and the 105 naming Windows 365. In those answers Windows 365 took the first choice thirty-two times and Amazon WorkSpaces two.
| Model | DirectAWW3 | ParaphraseAWW3 | ComparativeAWW3 | Budget-constrainedAWW3 | Scale-constrainedAWW3 | NegativeAWW3 |
|---|---|---|---|---|---|---|
| Claude Haiku 4.5 | AW | AW | ||||
| GPT-5.4 mini | AWW3 | AW | AW | AW | AW | AW |
| Gemini 3.5 Flash | W3 | AW | AWW3 | W3 | AW | |
| Perplexity Sonar | AW | |||||
| Grok 4.1 Fast | AW | AWW3 | AW | AWW3 | AW | AWW3 |
| Mistral Small | AW | AW | AWW3 | |||
| DeepSeek V4 Flash | AWW3 | AW | AWW3 | AWW3 | AWW3 | AW |
| Llama 4 Maverick | AW | |||||
| Qwen 3.7 Flash | AW | AW | AW | AWW3 | AW | |
| Kimi K2 | W3 | AWW3 | AWW3 | W3 | AWW3 | AWW3 |
| GLM 4.7 FlashX | AW | AWW3 | W3 | |||
| MiniMax M2.5 | AW | AW | AW | AW | AWW3 | |
| GPT-6 Luna | AWW3 | AWW3 | AWW3 | AWW3 | ||
| Muse Glimmer 30B | AW | AW | AW | AW |
Bold names in an answer are the products the judge labeled a first choice; a model naming several gives each of them that label. The full answer text for every row is in the record.
Every negative label with a quote, up to three, then the highest-weighted positives, up to three. Six of eight in this category shown.
“avoid or be cautious with Amazon WorkSpaces, Citrix DaaS, Azure Virtual Desktop (for non-Windows needs), and Shells” Mistral Small · negative prompt · hard negative
“Reasons to avoid: No text/data transfer from local machine to VM, additional technical support is paid” MiniMax M2.5 · negative prompt · hard negative
“AWS has published end-of-life timelines for some WorkSpaces client versions and support changes for PCoIP-based WorkSpaces Personal, which is a strong signal to avoid deployments that depend on deprecated clients or protocols.” GPT-5.4 mini · negative prompt · soft negative
“If I had to make a single recommendation without more constraints, I'd pick Amazon WorkSpaces for most mid-sized B2B companies” Perplexity Sonar · paraphrase prompt · first choice
“Windows 365 or Amazon WorkSpaces offer the best balance of affordability, reliability, and ease of management.” DeepSeek V4 Flash · budget prompt · first choice
“For most mid-sized B2B companies, Microsoft Azure Virtual Desktop or Amazon WorkSpaces are excellent choices” Mistral Small · paraphrase prompt · first choice
Every negative label with a quote, up to three, then the highest-weighted positives, up to three. Four of five in this category shown.
“Less flexible; per-user cost regardless of usage; not ideal for pooled/shared” DeepSeek V4 Flash · scale prompt · soft negative
“For most budget-constrained small businesses, Windows 365 or Amazon WorkSpaces offer the best balance of affordability, reliability, and ease of management.” DeepSeek V4 Flash · budget prompt · first choice
“For most small to midsize organizations, Windows 365 or Azure Virtual Desktop (if you are all-in on Windows and Azure) are simpler and more cost-effective choices.” Mistral Small · negative prompt · first choice
“Microsoft Windows 365 (Cloud PC): The easiest to manage if your IT team is already deeply integrated with Microsoft Intune.” Gemini 3.5 Flash · scale prompt · first choice
Comparisons are drawn for the top eight products in each category, each against each. The output is the models' output; nothing here is a recommendation by the index.