# Make vs Kestra: which do AI models recommend for orchestration, October 2026

IT AI Recommendation Index, October 2026 Edition, Workflow orchestration. Six of fourteen models named Make first on the direct prompt; zero named Kestra. Page: https://it-ai-index.com/it-data/workflow-orchestration/make-vs-kestra/

| | First-choice share | Rank | Negative rate | Labels | Models naming it |
|---|---|---|---|---|---|
| Make | 12% | #5 of 14 | 26% | 27 | 12 of 14 |
| Kestra | 3% | #7 of 14 | 11% | 19 | 8 of 14 |

## The direct prompt, model by model

- Claude Haiku 4.5: make first (first choices: Make, Zapier) (alternatives: Prefect, n8n)
- Gemini 3.5 Flash: make first (first choices: Make, Prefect) (alternatives: Astronomer Astro, Dagster, Workato, n8n)
- Grok 4.1 Fast: make first (first choices: Make, Zapier) (alternatives: Apache Airflow, Microsoft Power Automate, Prefect, Tray.io, n8n)
- Qwen 3.7 Flash: make first (first choices: Make) (alternatives: Kissflow, Microsoft Power Automate, Native HubSpot Workflows, Nintex)
- Kimi K2: make first (first choices: Make) (alternatives: Tray.io, Workato, n8n)
- GLM 4.7 FlashX: make first (first choices: Make) (alternatives: Zapier, n8n)
- Perplexity Sonar: neither first, one named (first choices: Workato) (alternatives: HubSpot workflows, Make, SnapLogic, Zapier, n8n)
- GPT-6 Luna: neither first, one named (first choices: Workato) (alternatives: Make, Prefect, Temporal, n8n)
- Muse Glimmer 30B: neither first, one named (first choices: Workato) (alternatives: Make, Microsoft Power Automate, ServiceNow Workflow, Zapier)
- GPT-5.4 mini: neither named (first choices: Prefect) (alternatives: Apache Airflow, Dagster, Temporal)
- Mistral Small: neither named (first choices: HubSpot Marketing Hub) (alternatives: Kissflow, Microsoft Power Automate, Revo, Tray.io)
- DeepSeek V4 Flash: neither named (first choices: n8n) (alternatives: Microsoft Power Automate, Zapier)
- Llama 4 Maverick: neither named (first choices: Kissflow) (alternatives: HubSpot Marketing Hub, Microsoft Power Automate)
- MiniMax M2.5: neither named (first choices: Microsoft Power Automate, Tray.io) (alternatives: Zapier)

## What the models said about Make

- "Avoid No-Code (Zapier/Make) for data infrastructure, ETL, and backend logic." (Gemini 3.5 Flash, negative prompt, hard negative)
- "If your budget is bleeding from expensive tools like Zapier or Make" (Gemini 3.5 Flash, budget prompt, hard negative)
- "Good visual builder at a budget price ($9/mo) — but workflows get messy at scale." (DeepSeek V4 Flash, direct prompt, soft negative)
- "Make is usually the best overall choice if you want a low-cost, user-friendly automation platform with a generous free tier" (Perplexity Sonar, budget prompt, first choice)
- "Make typically offers the best balance of features, pricing, and ease of use for mid-market B2B companies" (GLM 4.7 FlashX, direct prompt, first choice)
- "Make.com is the most cost-effective and powerful engine for a mid-market B2B budget" (Gemini 3.5 Flash, direct prompt, first choice)

## What the models said about Kestra

- "Kestra | Need mature ecosystem, extensive peer validation ... "younger ecosystem"" (Kimi K2, negative prompt, soft negative)
- "excellent alternatives to Airflow. However ... they carry a distinct business risk" (Gemini 3.5 Flash, negative prompt, soft negative)
- "Start with Prefect or Kestra for most teams - they offer the best balance of free licensing, modern features, and reasonable complexity." (GLM 4.7 FlashX, budget prompt, first choice)
- "n8n or Kestra offer the best balance of capability, ease of use, and zero licensing costs" (Kimi K2, budget prompt, first choice)
- "Kestra defines workflows in YAML rather than Python, making pipelines easier to version-control" (Claude Haiku 4.5, paraphrase prompt, alternative)

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. Comparisons are drawn for the top eight products in each category. Published under CC BY 4.0; the output is the models' output, and nothing here is a recommendation by the index.