# OpenMetadata vs Amundsen: which do AI models recommend for data catalogs, October 2026

IT AI Recommendation Index, October 2026 Edition, Data catalogs and governance. Zero of fourteen models named OpenMetadata first on the direct prompt; zero named Amundsen. Page: https://it-ai-index.com/it-data/data-catalogs-and-governance/openmetadata-vs-amundsen/

| | First-choice share | Rank | Negative rate | Labels | Models naming it |
|---|---|---|---|---|---|
| OpenMetadata | 18% | #2 of 15 | 7% | 27 | 13 of 14 |
| Amundsen | 4% | #7 of 15 | 33% | 12 | 7 of 14 |

## The direct prompt, model by model

- GPT-6 Luna: neither first, one named (first choices: Atlan) (alternatives: Microsoft Purview, OpenMetadata, Secoda)
- Claude Haiku 4.5: neither named (first choices: Atlan) (alternatives: Alation, OvalEdge)
- GPT-5.4 mini: neither named (first choices: DataHub) (alternatives: Alation, DataGalaxy)
- Gemini 3.5 Flash: neither named (first choices: Select Star) (alternatives: Atlan, Coalesce Catalog, OvalEdge, Secoda)
- Perplexity Sonar: neither named (first choices: Atlan) (alternatives: Alation, OvalEdge, Select Star)
- Grok 4.1 Fast: neither named (first choices: Atlan) (alternatives: Alation, OvalEdge, Secoda)
- Mistral Small: neither named (first choices: Atlan, OvalEdge)
- DeepSeek V4 Flash: neither named (first choices: Secoda) (alternatives: Atlan, DataHub)
- Llama 4 Maverick: neither named (first choices: Atlan)
- Qwen 3.7 Flash: neither named (first choices: Atlan) (alternatives: AWS DataZone, DataHub, Google Dataplex, OvalEdge, Seceda)
- Kimi K2: neither named (first choices: Secoda) (alternatives: Atlan, OvalEdge, Select Star)
- GLM 4.7 FlashX: neither named (first choices: Atlan, Secoda, Select Star) (alternatives: Informatica Intelligent Data Management Cloud, OvalEdge, data.world)
- MiniMax M2.5: neither named (first choices: Atlan, Secoda) (alternatives: Alation, Microsoft Purview)
- Muse Glimmer 30B: neither named (first choices: Atlan) (alternatives: Alation, OvalEdge)

## What the models said about OpenMetadata

- "While open source options like OpenMetadata and DataHub are credible, they require substantial in-house engineering expertise to deploy, maintain, and scale." (Mistral Small, negative prompt, soft negative)
- "Only choose this if you have a dedicated data platform engineer who has the bandwidth to manage and upgrade the infrastructure." (Gemini 3.5 Flash, scale prompt, soft negative)
- "OpenMetadata is generally the better starting point due to its simpler stack and faster time-to-value." (MiniMax M2.5, budget prompt, first choice)
- "OpenMetadata (Best Overall Open Source) ... this is widely considered the best open-source data catalog" (GLM 4.7 FlashX, budget prompt, first choice)
- "The Best All-Around (Open Source): OpenMetadata ... this is currently the top contender." (Qwen 3.7 Flash, budget prompt, first choice)

## What the models said about Amundsen

- "Avoid for Mid-Market: ... open-source like DataHub/Amundsen (free but needs engineering overhead)" (Grok 4.1 Fast, direct prompt, hard negative)
- "Amundsen (Lyft) — archived and no longer maintained as of 2026, so don't build on it" (DeepSeek V4 Flash, budget prompt, hard negative)
- "have lower license costs but require significant DevOps effort to install, maintain, and keep running" (Qwen 3.7 Flash, scale prompt, soft negative)
- "likely to be an open-source tool such as Amundsen or DataHub, which are suitable for on-premise environments" (Llama 4 Maverick, budget prompt, first choice)
- "Best ultra-lean choice: Amundsen" (GPT-5.4 mini, budget prompt, first choice)

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.