Aider Polyglot

At a glance

Methodology

225 exercism-style coding exercises in C++, Go, Java, JavaScript, Python and Rust. Score = pass_rate_2 (% of cases where the model's second attempt produces a passing solution).

Limitations

Aider's harness shapes prompts in a specific way — results may not directly compare to other coding benchmarks.

By provider

• OpenAI· 11 models

  88 % pass@2

  GPT-5
  Average: 53.62 % pass@2Best: 88 % pass@2
• Google· 1 model

  72.9 % pass@2

  Gemini 2.5 Pro
  Average: 72.9 % pass@2Best: 72.9 % pass@2
• Alibaba· 2 models

  59.6 % pass@2

  Qwen3 235B A22B