Appwrite Arena

Appwrite Arena

Compare AI models side by side with structured benchmarks covering coding, reasoning, and creative tasks — all focused on Appwrite.

appwrite/arena arena.appwrite.io
Appwrite Arena Thumbnail
TanStack Start React Tailwind CSS TypeScript

Minimal case study!

This is a small project made in a couple of days, and it only contains screenshots instead of a detailed case study. Some screenshots have captions to capture my thoughts and intentions when developing those parts of the application.

Hero section with leaderboards, ability to filter, sort, and multiple complementary metadata.
Hero section with leaderboards, ability to filter, sort, and multiple complementary metadata.
Same hero section toggled to card view for more vertical space coverage.
Same hero section toggled to card view for more vertical space coverage.
Just below hero section with raw data, a graph section for easier understanding and consumption of data.
Just below hero section with raw data, a graph section for easier understanding and consumption of data.
Ability to drill down into specific models of radar chart, for better visibility.
Ability to drill down into specific models of radar chart, for better visibility.
A graph showcasing how much smarter each model is when skill files are available through tools.
A graph showcasing how much smarter each model is when skill files are available through tools.
A bubble chart to quickly showcase how much value each model gives for its price.
A bubble chart to quickly showcase how much value each model gives for its price.
Graph to represent how large of a gap between MCQ ("simple") and Freeform ("complex") each model has.
Graph to represent how large of a gap between MCQ ("simple") and Freeform ("complex") each model has.
A final graph covering overall value of a model, comparing its intelligence to its price.
A final graph covering overall value of a model, comparing its intelligence to its price.
A section explaining why skills matter, with nice error and success logs for easier understanding of code samples. The section is followed by list of Appwrite services covered by this benchmark.
A section explaining why skills matter, with nice error and success logs for easier understanding of code samples. The section is followed by list of Appwrite services covered by this benchmark.
A final section explaining difference between MCQ and Freeform, a CTA to check out source code, and a footer.
A final section explaining difference between MCQ and Freeform, a CTA to check out source code, and a footer.
Detailed page for each model for SEO purposes, and ability to drill down into details. Contains additional metadata such as benchmark duration, price, and token usage. Below the table is a categorized set of all questions asked and model's answers, with quick indication of correctness.
Detailed page for each model for SEO purposes, and ability to drill down into details. Contains additional metadata such as benchmark duration, price, and token usage. Below the table is a categorized set of all questions asked and model's answers, with quick indication of correctness.
MCQ section shows a question, all options, and highlights the correct and wrong answer. It also includes exact prompt and output token usage, duration until the final answer, and the cost to get the answer.
MCQ section shows a question, all options, and highlights the correct and wrong answer. It also includes exact prompt and output token usage, duration until the final answer, and the cost to get the answer.
Freeform section presents a question, a sample correct answer, and a rubric for rating the model's answer. Below is the exact model's answer, and reasoning for the score given by an AI judge.
Freeform section presents a question, a sample correct answer, and a rubric for rating the model's answer. Below is the exact model's answer, and reasoning for the score given by an AI judge.