⌕Is ___ SOTA?
benchmarks / aider-polyglot

Aider Polyglot

percent correct. Status active. Contamination risk med. Maintainer Aider / Epoch collection. Source.

ModelScoreDateKind
GPT Oss 120B41.82025-08-05observation
Qwen3 235B A22b59.62025-04-29observation
DeepSeek V3.2 Exp70.22025-09-29observation
DeepSeek V3.2 Exp Thinking74.22025-09-29observation
GPT-581.32025-08-25observation
GPT-586.72025-08-25observation
GPT-5882025-08-23observation
O376.92025-04-16observation
Grok 3 Mini Beta34.72025-04-10observation
Openhands LM 32B V0.110.22025-04-19observation
Kimi K2 Instruct59.12025-07-12observation
Grok 4 070979.62025-07-09observation
O3 Pro84.92025-06-28observation
Gemini 2.5 Pro Preview 05.0676.92025-05-07observation
Gemini 2.5 Pro Preview 06.0579.12025-06-06observation
Gemini 2.5 Flash Preview 05.20442025-05-26observation
Claude Sonnet 456.42025-05-24observation
Claude Sonnet 4 2025051461.32025-05-24observation
Claude Opus 470.72025-05-25observation
Claude Opus 4 20250514722025-05-25observation
Qwen3 32B402025-05-08observation
Llama 4 Maverick 17B 128e Instruct15.62025-04-06observation
GPT-4.1 Mini32.42025-04-14observation
GPT-4.1 Nano8.92025-04-14observation
ChatGPT 4O 03.27 202545.32025-03-29observation
Gemini 2.5 Flash Preview 04.1747.12025-04-20observation
Grok 3 Mini Beta49.32025-04-10observation
GPT-4.152.42025-04-14observation
Grok 3 Beta53.32025-04-10observation
DeepSeek V3 032455.12025-03-24observation