Codeforces — Same Codeforces percentile metric; both from DeepSeek-R1 card family, ranges/models align
| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | DeepSeek-V4-Pro | DeepSeek | 3206 | 5 | 2026-06-27 |
| 2 | GPT-5.4 | OpenAI | 3168 | 3 | 2026-07-04 |
| 3 | DeepSeek-V4-Flash | DeepSeek | 3052 | 2 | 2026-06-27 |
| 4 | Gemini 3.1 Pro | 3052 | 4 | 2026-07-04 | |
| 5 | o1 | OpenAI | 2061 | 5 | 2026-06-04 |
| 6 | DeepSeek-R1 | DeepSeek | 2029 | 6 | 2026-06-04 |
| 7 | OpenAI o1-mini | OpenAI | 1820 | 5 | 2026-06-04 |
| 8 | DeepSeek-R1-Distill-Qwen-32B | DeepSeek | 1691 | 4 | 2026-06-04 |
| 9 | DeepSeek-R1-Distill-Llama-70B | DeepSeek | 1633 | 4 | 2026-06-04 |
| 10 | DeepSeek-R1-Distill-Qwen-14B | DeepSeek | 1481 | 4 | 2026-06-04 |
| 11 | R1-Zero | DeepSeek | 1444 | 1 | 2026-05-03 |
| 12 | QwQ 32B Preview | Alibaba | 1316 | 4 | 2026-06-04 |
| 13 | DeepSeek-R1-Distill-Llama-8B | DeepSeek | 1205 | 4 | 2026-06-04 |
| 14 | DeepSeek-R1-Distill-Qwen-7B | DeepSeek | 1189 | 4 | 2026-06-04 |
| 15 | DeepSeek-V3 | DeepSeek | 1134 | 6 | 2026-06-04 |
| 16 | DeepSeek-R1-Distill-Qwen-1.5B | DeepSeek | 954 | 4 | 2026-06-04 |
| 17 | GPT-4o | OpenAI | 759 | 6 | 2026-06-04 |
| 18 | Claude 3.5 Sonnet | Anthropic | 717 | 6 | 2026-06-04 |
| 19 | DeepSeek-V2.5 | DeepSeek | 35.6 | 1 | 2026-05-03 |
| 20 | Llama 3.1 Instruct 405B | Meta | 25.3 | 1 | 2026-05-03 |
| 21 | Qwen2.5 Instruct 72B | Alibaba | 24.8 | 1 | 2026-05-03 |
| 22 | DeepSeek-V2 | DeepSeek | 17.5 | 1 | 2026-05-03 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.