Zero of fourteen models named Amazon ECR first on the direct prompt; one named GitHub Packages. Amazon ECR was named by eight of the fourteen models and GitHub Packages by twelve and Amazon ECR carries 12 labels and GitHub Packages 42, so the shares are not directly comparable.
Named in two categories 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 artifact and container registries page.
| Model | DirectAEGP | ParaphraseAEGP | ComparativeAEGP | Budget-constrainedAEGP | Scale-constrainedAEGP | NegativeAEGP |
|---|---|---|---|---|---|---|
| Claude Haiku 4.5 | AE | GP | ||||
| GPT-5.4 mini | AE | |||||
| Gemini 3.5 Flash | GP | GP | GP | GP | GP | |
| Perplexity Sonar | AE | |||||
| Grok 4.1 Fast | GP | GP | GP | |||
| Mistral Small | GP | GP | ||||
| DeepSeek V4 Flash | GP | GP | GP | |||
| Llama 4 Maverick | ||||||
| Qwen 3.7 Flash | GP | GP | ||||
| Kimi K2 | GP | GP | GP | GP | ||
| GLM 4.7 FlashX | AE | AEGP | GP | |||
| MiniMax M2.5 | GP | GP | ||||
| GPT-6 Luna | AE | GP | GP | |||
| Muse Glimmer 30B | AE | GP |
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. Three of four in this category shown.
“If you want a single default recommendation without more context, I'd lean toward Amazon ECR for a mid-sized B2B company” GPT-5.4 mini · paraphrase prompt · first choice
“AWS Elastic Container Registry (ECR) is the obvious choice if you're running on AWS” Claude Haiku 4.5 · paraphrase prompt · first choice
“For most mid-sized B2B companies, I'd start with Amazon ECR if you're on AWS” GLM 4.7 FlashX · paraphrase 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.
“avoid direct, unrestricted use of public community images from Docker Hub, GitHub Container Registry, npm/PyPI, Artifact Hub” Muse Glimmer 30B · negative prompt · hard negative
“GitHub Packages does not natively support PyPI (Python). If your team writes a significant amount of Python, GitHub Packages will not be a complete solution.” Gemini 3.5 Flash · direct prompt · soft negative
“Cons: Often less mature in terms of advanced proxying, virtual repositories, and advanced binary management compared to dedicated tools.” Gemini 3.5 Flash · scale prompt · soft negative
“For most small to medium companies, GitHub Packages is often the best choice if you're already using GitHub” Claude Haiku 4.5 · budget prompt · first choice
“The Best "All-in-One" Option: GitHub Packages / GitLab Package Registry” Gemini 3.5 Flash · budget prompt · first choice
“Top Recommendation: GitHub Packages” Kimi K2 · 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.