BigCodeArena

All projects

Apps & Tools

An execution-based evaluation platform for interacting with and assessing AI-generated code.

byCody Horvath

Playable

BigCodeArena media is blocked

Allow external media to connect to the provider and play this content.

Description

BigCodeArena is a human-in-the-loop platform designed to evaluate code functionality and quality across various programming languages through live execution. Unlike static text-based benchmarks, it features a showcase of interactive applications such as physics simulations (Black Hole Orbital Simulation, Ocean Dynamics Lab) and UI utilities that users can run to assess performance.

Descriptions, tags, and model credits may be AI-generated or inferred from public sources and can be incomplete or wrong. Learn more.

Details

Why this is interesting: It introduces a novel execution-based human evaluation layer for code models, moving beyond simple text-based…
Attributed to: Cody Horvath@bigcode
Built with: GPT-5 DeepSeek V3.2 Qwen3.6
Links: huggingface.co
View tweet
Tags: #coding #evaluation #benchmark #simulation

Discovered: X
First posted: X · Oct 8, 2025
Listed: Added 50d ago
Build: Strong
Models: Strong

Prompt

Given me a Pygame of a ball bouncing inside a spinning hexagon, with the full control

Source

BigCodeArena

Description

Timeline

BigCodeArena

Description

Timeline