October 2026 Edition, published · The permanent record of that edition, superseded since. The latest edition →
Index › HR › Editions › October 2026 Edition
October 2026 Edition

October 2026 Edition

Published . This page is the permanent record of the October 2026 Edition: the standing of every category as measured in it, under vendor table v2026-11.3. The unqualified category addresses and the home page carry the latest edition.

58 categories · 14 models · 348 prompts · 4,872 answers · 25,742 judge labels, all published

The leaderboard

The first choice in all 58 categories · 51 unsettled
The latest edition →
Share is the count of first choices across the direct, paraphrase, budget and scale prompts, over all fourteen models. Each row opens that category's page for this edition.What the verdicts mean
Workforce planning and analytics
Employee monitoring
A different leader in every segment
Compensation and total rewards
HSA and FSA administrators
Mid-market and enterprise share a leader. Small business differs.
Core HR and payroll
Whistleblowing and investigations
Small business and mid-market share a leader. Enterprise differs.
Compensation and total rewards
Financial wellness platforms
Mid-market and enterprise share a leader. Small business differs.
Core HR and payroll
PEO providers
Small business and mid-market share a leader. Enterprise differs.
Onboarding and employee experience
Employee giving and volunteering
A different leader in every segment
Talent acquisition
Staffing agency platforms
Small business and mid-market share a leader. Enterprise differs.
Compensation and total rewards
401(k) providers
A different leader in every segment

Movement since September 2026

Between the September 2026 Edition and the October 2026 Edition the leader changed in twelve of fifty categories, four of them by more than the noise floor of 13 points. Fourteen of 498 changes in first-choice share clear the floor. Movement is read over the twelve models both editions asked; GPT-6 Luna, Muse Glimmer 30B joined this edition and are in the standing but not yet in the comparison.
50Categories compared
498Products changed
14Moved past the floor
12Leader changes
6Verdict changes

Leader changes, mid-market buyers · 12, of which 4 cleared the floor

GoPerfect→Manatal33%
New leader · moved

31 points clear of GoPerfect's current 2%.

PlanSource→ADP Workforce Now21%
New leader · moved

16 points clear of PlanSource's current 5%.

Names swapped within the floor · 8, read as unsettled

Verdict changes{for_seg}

Biggest movers in first-choice share, mid-market buyers

−3135% → 4%
MovedCleared the floor by 18 points · rank 1 → 9 of 110
+2112% → 33%
MovedCleared the floor by 8 points · rank 3 → 1 of 84
+1926% → 45%
MovedCleared the floor by 6 points · rank 1 of 93, unchanged
+192% → 21%
MovedCleared the floor by 6 points · rank 6 → 1 of 60
−1868% → 50%
MovedCleared the floor by 5 points · rank 1 of 115, unchanged
−1823% → 5%
MovedCleared the floor by 5 points · rank 1 → 6 of 60
+1832% → 50%
MovedCleared the floor by 5 points · rank 1 of 89, unchanged
−1728% → 11%
MovedCleared the floor by 4 points · rank 1 → 5 of 83
−1719% → 2%
MovedCleared the floor by 4 points · rank 1 → 10 of 84
−1616% → 0%
MovedCleared the floor by 3 points · rank 1 → 19 of 83
−1622% → 6%
MovedCleared the floor by 3 points · rank 2 → 4 of 117
−1517% → 2%
MovedCleared the floor by 2 points · rank 3 → 8 of 55
0Floor 13 pts · measured

The floor is 13 points of share, measured: how far the models move a leader on their own when the same questions are asked twice with nothing changed. A larger change is movement; a smaller one can't be told apart from repeat variation, and both are shown. Each index measures its own floor from its own repeat, over its own categories, so the floor differs from one index to the next. Movement is read over the twelve models both editions asked; GPT-6 Luna, Muse Glimmer 30B joined this edition and are in the standing but not yet in the comparison. How the floor is measured · Every change resolves to its category page.

The models asked

Fourteen models: the generally available models a buyer meets as a default in an assistant, with search on. Two joined this edition and zero left. A model joins at an edition boundary and stays at least three editions; movement is read over the models both editions asked, so a join never moves a share on its own. The roster rule.
ModelVersion asked forLabHow it searchesIn the set since
Claude Haiku 4.5claude-haiku-4-5Anthropicnative searchOctober 2026
GPT-5.4 minigpt-5.4-miniOpenAInative searchOctober 2026
Gemini 3.5 Flashgemini-3.5-flashGooglenative searchOctober 2026
Perplexity SonarsonarPerplexitynative searchOctober 2026
Grok 4.20spacexai/grok-4.20-non-reasoningxAIgateway, Perplexity searchNovember 2026
Mistral Smallmistral/mistral-smallMistralgateway, Perplexity searchOctober 2026
DeepSeek V4 Flashdeepseek/deepseek-v4-flashDeepSeekgateway, Perplexity searchOctober 2026
Llama 4 Maverickmeta/llama-4-maverickMetagateway, Perplexity searchOctober 2026
Qwen 3.7 Flashalibaba/qwen3.7-flashAlibabagateway, Perplexity searchOctober 2026
Kimi K2moonshotai/kimi-k2Moonshotgateway, Perplexity searchOctober 2026
GLM 4.7 FlashXzai/glm-4.7-flashxZ.AIgateway, Perplexity searchOctober 2026
MiniMax M2.5minimax/minimax-m2.5MiniMaxgateway, Perplexity searchOctober 2026
GPT-6 Lunagpt-6-lunaOpenAInative searchOctober 2026 · joined this edition
Muse Glimmer 30Bmeta/muse-glimmer-30bMetagateway, Perplexity searchOctober 2026 · joined this edition
← September 2026 EditionAll editionsNovember 2026 Edition →