Deepseek R1 Distill Llama 70B is an LLM that excels in math, code, and reasoning tasks, achieving performance comparable to frontier models like OpenAI-o1 through advanced distillation techniques based on Llama-3.3-70B-Instruct.

Some other noteworthy features of Deepseek R1 Distill Llama 70B include strong performance on mathematical benchmarks with 94.5% pass rate on MATH-500 and 70.0% on AIME 2024, as well as competitive coding abilities with a CodeForces Rating of 1633.

Metric	Value
Parameter Count	70 billion
Mixture of Experts	Yes
Active Parameter Count	37 billion
Context Length	128K-131K tokens
Multilingual	Yes
Quantized*	Unknown

*Quantization is specific to the inference provider and the model may be offered with different quantization levels by other providers.