Coding

Aider Polyglot

Multi-language coding agent benchmark: edit existing codebases to pass unit tests across languages; score depends on the Aider model prompt and edit strategy.

How models are tested

Scores are published numbers from the sources below. They are only comparable when the evaluation version, tools, agent harness, and sampling setup match. Missing scores are omitted rather than treated as zero.

Category: Coding
Metric: Percent
Direction: Higher is better
Coverage: 35 / 318Models in this catalog with a published score
Updated: 2026-09-29Catalog snapshot date
Source: aider.chat/docs/leaderboards/

Scores

1
GPT-5
OpenAIclosed
88.0%
2
o3-pro
OpenAIclosed
84.9%
3
o3
OpenAIclosed
81.3%
4
Grok 4
xAIclosed
79.6%
5
Gemini 2.5 Pro
Googleclosed
79.1%
6
DeepSeek V3.1
DeepSeekopen
76.3%
7
GPT-5 Mini
OpenAIclosed
71.6%
8
DeepSeek R1-0528
DeepSeekopen
71.4%
9
Claude Opus 4
Anthropicclosed
70.7%
10
o1
OpenAIclosed
61.7%
11
Claude 3.7 Sonnet
Anthropicclosed
60.4%
12
Qwen3 235B-A22B
Alibabaopen
59.6%
13
Kimi K2
Moonshotopen
59.1%
14
o4-mini
OpenAIclosed
58.2%
15
DeepSeek R1
DeepSeekopen
56.9%
16
Claude Sonnet 4
Anthropicclosed
56.4%
17
DeepSeek V3
DeepSeekopen
55.1%
18
o3-mini
OpenAIclosed
53.8%
19
Grok 3
xAIclosed
53.3%
20
GPT-4.1
OpenAIclosed
52.4%
21
Claude 3.5 Sonnet (Oct)
Anthropicclosed
51.6%
22
Qwen3-Next 80B-A3B
Alibabaopen
49.8%
23
GPT-5 Nano
OpenAIclosed
48.4%
24
GPT-4.5
OpenAIclosed
44.9%
25
gpt-oss-120B
OpenAIopen
41.8%
26
Qwen3 32B
Alibabaopen
40.0%
27
gpt-oss-20B
OpenAIopen
34.2%
28
o1 Mini
OpenAIclosed
32.9%
29
GPT-4.1 Mini
OpenAIclosed
32.4%
30
QwQ-32B
Alibabaopen
20.9%
31
Llama 4 Maverick
Metaopen
15.6%
32
Yi Lightning
01.AIclosed
12.9%
33
GPT-4.1 Nano
OpenAIclosed
8.9%
34
Gemma 3 27B
Googleopen
4.9%
35
GPT-4o Mini
OpenAIclosed
3.6%

Back to all benchmarks