MDL
DeepSeek-R1-Distill-Llama-70B
deepseek-ai
Parameters
70B
Downloads
110.5K
on HF
Likes
776
Avg score
27.8
6 benchmarks
Weights
Open
Released
Jan 20, 2025
About
DeepSeek-R1-Distill-Llama-70B is a 70B-parameter open-weights model in the DeepSeek family from deepseek-ai. Benchmarked on 6 evals, averaging 27.8.
mittext-generation
Benchmark scores · 6
avg 27.8 — higher is betterThe research behind this model
all papers mentioning it →DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement LearningPlans Don't Persist: Why Context Management Is Load Bearing for LLM Agents▲ 1 on HF · Jun 22, 2026Theory-optimal Quantization Based on FlatnessMay 11, 2026
Mentions matched by name in title/abstract — from the arXiv + HF daily corpus.
History
full ledger →Capability over time
Append-only ledger — every observed change to this model's numbers.