Home Model series DeepSeek R1 DeepSeek-R1 DeepSeek-R1 · AMC 12 · 2024; DeepSeek-R1 paper Table 13 · Scores and configurations DeepSeek-R1 · AMC 12 · 2024; DeepSeek-R1 paper Table 13 · Scores and configurations DeepSeek-R1 record for AMC 12 · 2024; DeepSeek-R1 paper Table 13. 1 score record for this release, including 1 exact value (may include historical records). Includes 1 source location.
Scores and evaluation conditions DeepSeek-R1 · AMC 12 scores and configurations Metric and score Configuration and evaluation conditions Sources Points 11.3 points/15 Reported 11.3/15 · Unit points/15 Published Unconfirmed
DeepSeek-R1 reported reasoning configuration Settings Maximum generation length 32,768 tokens. For sampled benchmarks: temperature=0.6, top_p=0.95, 64 responses per query to estimate Pass@1. Unknowns Per-benchmark sampling use is not itemized beyond the card preface. Tools Tool access not recorded Reasoning budget Output limit 32768 tokens Framework not recorded Protocol DeepSeek-R1 paper Table 13; AIME raw score shown as points out of 15. Conditions DeepSeek R1 row. View this score in rankings
Benchmark versions and metrics Points · points/15 · Higher is better