Tree-sitter allows aider to build a repo map that better summarizes large code bases.| aider
Benchmark results for Qwen3 models using the Aider polyglot coding benchmark.| aider
The $6.32 benchmark cost reported for Gemini 2.5 Pro Preview 03-25 was incorrect.| aider
DeepSeek's API has been experiencing reliability issues. Here are alternative providers you can use.| aider
R1+Sonnet has set a new SOTA on the aider polyglot benchmark. At 14X less cost compared to o1.| aider
Reliably packaging & distributing python CLI tools is hard. Aider uses uv in novel ways to make it easy to install the aider CLI, its dependencies and python 3.12. All in an isolated env.| aider
QwQ is reasoning model like o1, and needs to be used as an architect with another model as editor.| aider
Open source LLMs are becoming very powerful, but pay attention to how you (or your provider) are serving the model. It can affect code editing skill.| aider
An Architect model describes how to solve the coding problem, and an Editor model translates that into file edits. This Architect/Editor approach produces SOTA benchmark results.| aider
Preliminary benchmark results for the new OpenAI o1 models.| aider
Sonnet’s score on the aider code editing benchmark has been stable since it launched.| aider
LLMs write worse code if you ask them to return the code wrapped in JSON via a tool function call.| aider
o1 scores the top result on aider’s new multi-language, more challenging coding benchmark.| aider