LLM2014 Code 2025-09 - C#: leaderboard
LLM2014 monthly coding benchmark for 2025-09, measuring C# code-generation problem-solving on the source score scale.
Metric: Score. Source: raw.githubusercontent.com. Saturation forecast: Estimated already saturated. 20 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek V3.1 (Thinking) | 8.7 | #264 (DeepSeek V3.1) |
| 2 | O4 Mini (High) | 8.65 | #180 (O4 Mini) |
| 3 | Claude Sonnet 4 (Thinking) | 8.31 | #199 (Claude Sonnet 4) |
| 4 | GPT-5 Mini (High) | 8.18 | #186 (GPT-5 Mini) |
| 5 | Gemini 2.5 Flash (Thinking) | 7.98 | #245 (Gemini 2.5 Flash) |
| 6 | Claude Sonnet 4 | 7.91 | #199 |
| 7 | Gemini 2.5 Pro | 7.15 | #152 |
| 8 | Qwen 3 Coder 480B A35B | 7.03 | #279 |
| 9 | GLM-4.5 (Thinking) | 6.94 | #275 (GLM-4.5) |
| 10 | Grok 4 | 6.8 | #181 |
| 11 | Seed-1.6 (Thinking) | 6.31 | #265 (Seed-1.6) |
| 12 | GPT-OSS-120B (High) | 6.27 | #341 (GPT-OSS-120B) |
| 13 | DeepSeek V3.1 | 6.2 | #264 |
| 14 | GPT-OSS-20B (High) | 5.89 | #518 (GPT-OSS-20B) |
| 15 | Qwen 3 235B A22B 2507 Instruct | 5.7 | #300 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=llm2014-code-2025-09-c-sharp · How It Works · Data refreshed daily, snapshot 2026-10-20.