Millie vs Percent Pledge
Four of fourteen models named Millie first on the direct prompt; two named Percent Pledge. Millie was named by thirteen of the fourteen models and Percent Pledge by ten and Millie carries 44 labels and Percent Pledge 21, so the shares are not directly comparable.
Millie
accepted challengerNamed in one category this edition.
Named in one category 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 employee giving and volunteering page.
By framing
Across every category in the November 2026 Edition, Millie and Percent Pledge were named in the same answer twenty-eight times, of the 108 answers naming Millie and the 47 naming Percent Pledge. In those answers Percent Pledge took the first choice three times and Millie thirteen.
Every model, every framing
| Model | DirectMI | ParaphraseMI | ComparativeMI | Budget-constrainedMI | Scale-constrainedMI | NegativeMI |
|---|---|---|---|---|---|---|
| Claude Haiku 4.5 | MI | MI | MI | MI | ||
| GPT-5.4 mini | MI | MI | ||||
| Gemini 3.5 Flash | | MI | MI | MI | | |
| Perplexity Sonar | MI | MI | MI | MI | ||
| Grok 4.20 | MI | | MI | MI | ||
| Mistral Small | MI | MI | MI | MI | ||
| DeepSeek V4 Flash | MI | MI | MI | |||
| Llama 4 Maverick | | |||||
| Qwen 3.7 Flash | MI | | | MI | ||
| Kimi K2 | MI | MI | MI | | ||
| GLM 4.7 FlashX | MI | MI | ||||
| MiniMax M2.5 | MI | MI | ||||
| GPT-6 Luna | MI | MI | MI | MI | MI | |
| Muse Glimmer 30B | MI | MI |
Percent PledgeMI first choiceMI named as an alternativeMI argued againstblank: not namedEach cell is one answer, Millie on the left and Percent Pledge on the right.The direct prompt
Millie first, Percent Pledge not the choice
4 of 14 modelsPercent Pledge was named in the answer but not as the choice, or not at all.Percent Pledge first, Millie an alternative
2 of 14 modelsMillie was named in the answer but not as the choice, or not at all.Neither was the first choice, one was named
5 of 14 modelsThe answer put something else first and named one of the two as an alternative.Neither was named
3 of 14 modelsThe answer made no first choice from these two in this category.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.
By buyer segment
Percent Pledge3%#6 of 11Percent Pledge3%#7 of 12Percent Pledge0%#– of 10What the models said about Millie
Every negative label with a quote, up to three, then the highest-weighted positives, up to three. Five of six in this category shown.
“These may be unsuitable where native nonprofit-CRM integration or federated/community-campaign support is essential” Perplexity Sonar · negative prompt · soft negative
“Groundswell and Millie — ask what screening they use now.” GPT-6 Luna · negative prompt · soft negative
“Millie supports donation matching programs at any budget level with customizable match ratios and caps” Claude Haiku 4.5 · budget prompt · first choice
“Uncommon Giving and Millie are the top recommendations for companies with limited budgets” Mistral Small · budget prompt · first choice
“the strongest default recommendation is Millie (formerly Groundswell)” Perplexity Sonar · paraphrase prompt · first choice
What the models said about Percent Pledge
Every negative label with a quote, up to three, then the highest-weighted positives, up to three. Three of four in this category shown.
“The best employee giving and volunteering platform for a mid-market B2B company is Percent Pledge, Uncommon Giving, or Selflessly.” Llama 4 Maverick · direct prompt · first choice
“lean toward modern, high-agility CSR tools (such as *Deed*, *Groundswell*, *Percent Pledge*, or *Uncommon Giving*)” Gemini 3.5 Flash · negative prompt · first choice
“Percent Pledge (Best Overall for Mid-Market)” MiniMax M2.5 · direct 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.