Zero of fourteen models named IBM Process Mining first on the direct prompt; two named Celonis Process Intelligence Platform. Both were named by all fourteen models and IBM Process Mining carries 37 labels and Celonis Process Intelligence Platform 64, so the shares are not directly comparable.
Named 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 process mining page.
Across every category in the October 2026 Edition, IBM Process Mining and Celonis Process Intelligence Platform were named in the same answer 100 times, of the 107 answers naming IBM Process Mining and the 191 naming Celonis Process Intelligence Platform. In those answers Celonis Process Intelligence Platform took the first choice thirty-four times and IBM Process Mining ten.
| Model | DirectIP | ParaphraseIP | ComparativeIP | Budget-constrainedIP | Scale-constrainedIP | NegativeIP |
|---|---|---|---|---|---|---|
| Claude Haiku 4.5 | IP | IP | ||||
| GPT-5.4 mini | IP | |||||
| Gemini 3.5 Flash | IP | IP | IP | |||
| Perplexity Sonar | IP | |||||
| Grok 4.1 Fast | IP | IP | IP | |||
| Mistral Small | IP | IP | ||||
| DeepSeek V4 Flash | IP | IP | IP | IP | ||
| Llama 4 Maverick | IP | |||||
| Qwen 3.7 Flash | IP | IP | IP | |||
| Kimi K2 | IP | IP | IP | |||
| GLM 4.7 FlashX | IP | |||||
| MiniMax M2.5 | IP | IP | IP | |||
| GPT-6 Luna | ||||||
| Muse Glimmer 30B | IP | IP | IP |
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.
“What I would avoid on a limited budget: Celonis, IBM Process Mining, and similar enterprise suites” GPT-5.4 mini · budget prompt · hard negative
“~$4,250/month and enterprise-governance-focused — overkill for most mid-market teams.” DeepSeek V4 Flash · direct prompt · hard negative
“IBM Process Mining starts at ~$4,250/month — still expensive for SMBs” DeepSeek V4 Flash · budget prompt · hard negative
“I would recommend IBM Process Mining if you want a process intelligence platform that is explicitly described as *tailored more to midsize organizations*” Perplexity Sonar · paraphrase prompt · first choice
“For most mid-sized B2B companies, IBM Process Mining or SAP Signavio are the best recommendations.” Mistral Small · paraphrase prompt · first choice
“Mid-Market & Agile Specialized Tools (e.g., Apromore, Fluxicon Disco, IBM Process Mining)” Gemini 3.5 Flash · scale prompt · first choice
Every negative label with a quote, up to three, then the highest-weighted positives, up to three. Six of eight in this category shown.
“What to Avoid (Unless You Have the Budget) ... Celonis | $60K–$100K/year + $40K–$80K implementation | Case-volume pricing adds up fast; opaque enterprise sales” Kimi K2 · direct prompt · hard negative
“What I would avoid on a limited budget: Celonis, IBM Process Mining, and similar enterprise suites” GPT-5.4 mini · budget prompt · hard negative
“avoid enterprise-first tools like Celonis, which are often described as more expensive” Perplexity Sonar · budget prompt · hard negative
“The undisputed heavyweight market leader... If you have a highly complex, multi-vendor IT architecture and a large budget... Choose Celonis.” Gemini 3.5 Flash · comparative prompt · first choice
“Celonis — The Market Leader ... Key Differentiator: Deep execution management with AI-powered recommendations and the broadest ecosystem” Kimi K2 · comparative prompt · first choice
“Celonis remains the undisputed heavyweight, commanding a 31.2% market share through its sophisticated Execution Management System (EMS)” Claude Haiku 4.5 · comparative 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.