According to CoinMeta, DHH (the author of Ruby, on, and Rails) conducted a programming model test, comparing multiple models using the task of converting Python to Rust. DeepSeek Pro V4 Max was ultimately completed in about 2.5 hours, with a total token cost of $23. Grok took 55 dollars and about 1.5 hours, while GPT-5.6 and SOL cost $43 each. DeepSeek V4 Flash and GPT-5.6 and Luna failed to complete the task. The initial round was led by Fable, and DHH estimated the cost at around $550, taking about 45 minutes. The project has been made open-source, and the repository contains 354 consistency tests that compare frame-by-frame and byte-by-byte the Rust version with the original Python version. The final product can be compiled into a static single file of about 3.3MB.