← Benchmarks
CodeForces
A competitive programming benchmark using problems from the CodeForces platform. The benchmark evaluates code generation capabilities of LLMs on algorithmic problems with difficulty ratings ranging from 800 to 2400. Problems cover diverse algorithmic categories including dynamic programming, graph algorithms, data structures, and mathematical problems with standardized evaluation through direct platform submission.
id codeforces · max 3000 · 20 models reported
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek-V4.1-Flash DeepSeek · open | 3471.00 |
| 2 | Qwen3 14B Alibaba Cloud / Qwen Team · open | 1766.00 |
| 3 | Seed 2.0 Mini ByteDance | 1644.00 |
| 4 | DeepSeek-V4-Flash-Max DeepSeek · open | 1.00 |
| 5 | DeepSeek-V4-Pro-Max DeepSeek · open | 1.00 |
| 6 | DeepSeek-V4-Flash-0423 DeepSeek · open | 0.94 |
| 7 | DeepSeek-V3.2-Speciale DeepSeek · open | 0.90 |