DeepSeek R1 Distill Llama 70B

Overview

DeepSeek-R1 is the first-generation reasoning model built atop DeepSeek-V3 (671B total parameters, 37B activated per token). It incorporates large-scale reinforcement learning (RL) to enhance its chain-of-thought and reasoning capabilities, delivering strong performance in math, code, and multi-step reasoning tasks.

DeepSeek R1 Distill Llama 70B was released on January 20, 2025. API access is available through DeepInfra.

Benchmarks

DeepSeek R1 Distill Llama 70B Performance Across Datasets

Scores sourced from the model's scorecard, paper, or official blog posts

llm-stats.com - Mon Dec 22 2025

Pricing

Pricing, performance, and capabilities for DeepSeek R1 Distill Llama 70B across different providers:

Provider	Input ($/M)	Output ($/M)	Max Input	Max Output	Latency (s)	Throughput	Quantization	Input	Output
DeepInfra	$0.10	$0.40	128.0K	128.0K	0.65	37.0 tok/s	—	Text Image Audio Video	Text Image Audio Video

API Access