DeepSeek R1 Distilled Llama 70B
Llama 70B distilled with reasoning capabilities from Deepseek R1. Surpasses GPT-4o with 94.5% on MATH-500 & matches o1-mini on coding.
About model
DeepSeek R1 Distilled Llama 70B performs complex reasoning tasks, excelling in math, code, and reasoning benchmarks. It is designed for researchers and developers seeking advanced language models.
Quickstart guides
Performance benchmarks
Model | AIME 2025 | GPQA Diamond | HLE | LiveCodeBench | MATH500 | SWE-bench verified |
|---|---|---|---|---|---|---|
DeepSeek R1 Distilled Llama 70B | 65.0% | 78.3% | Related open-source models | Competitor closed-source models | ||
 Claude Opus 4.6 | 90.5% | 34.2% | 78.7% | |||
 OpenAI o3 | 83.3% | 24.9% | 99.2% | 62.3% | ||
OpenAI o1 | 76.8% | 96.4% | 48.9% | |||
GPT-4o | 49.2% | 2.7% | 32.3% | 89.3% | 31.0% |
Related models
- TypeChatReasoning
- Main use casesChatReasoning
- FeaturesJSON Mode
- Fine tuningSupported
- DeploymentOn-Demand DedicatedMonthly Reserved
- Parameters70B
- Context length128K
- Input price
$2.00 / 1M tokens
- Output price
$2.00 / 1M tokens
- Input modalitiesText
- Output modalitiesText
- ReleasedJanuary 20, 2025
- Last updatedDecember 22, 2025
- Quantization levelFP16
- External link
- CategoryChat