DeepSeek: R1 Distill Llama 70B
Analysis Summary
R1 Distill Llama 70B is a distilled reasoning model combining DeepSeek's R1 training with Meta's Llama 70B architecture. Its math index of 53.7 and AIME-25 score of 0.537 reflect genuine mathematical strength. MMLU-Pro of 0.795 is solid for a distilled model. However, LiveCodeBench of 0.266 and terminal benchmark scores are weak, limiting coding utility.
For businesses, this model suits mathematical reasoning, structured analysis, and moderate-complexity content tasks. Its agentic index of 11.7 is low, and long-context retrieval is poor (0.11), making it unsuitable for document-heavy or multi-step agent workflows. A -4 point regional penalty applies.
At $0.80 per million tokens for both input and output, it offers reasonable value for math-focused use cases. Teams needing affordable mathematical reasoning without frontier-level coding or agentic capability will find it a practical fit.
Assessed June 30, 2026
Editorial notes
DeepSeek R1 Distill Llama 70B shows strong math performance and reasonable MMLU-Pro scores at competitive pricing, but coding and agentic benchmarks are limited for business-critical workflows.
Rankings consider pricing, capabilities, benchmarks, and real-world applicability and are refreshed as new models launch. Feedback?
Performance Profile
How DeepSeek: R1 Distill Llama 70B compares
DeepSeek: R1 Distill Llama 70B ranks #233 of 385 AI models we track for overall intelligence, #249 of 293 for agentic tasks. Its 128K-token context window is larger than 43% of the models we list. At $0.80 per million input tokens it is cheaper than 28% of comparable models.
About DeepSeek: R1 Distill Llama 70B
DeepSeek R1 Distill Llama 70B is a distilled large language model based on Llama-3.3-70B-Instruct, using outputs from DeepSeek R1. The model combines advanced distillation techniques to achieve high performance across..
Architecture Detail
| Instruct Type | deepseek-r1 |
Performance Indices
Source: Artificial Analysis
Benchmark Scores
Intelligence
Technical
Content
Benchmark data from Artificial Analysis and Hugging Face
How does DeepSeek: R1 Distill Llama 70B stack up?
Compare side-by-side with other professional models.
Model Information
Pricing
| Token Type | Cost per 1M tokens | Cost per 1K tokens |
|---|---|---|
| Input | $0.80 | $0.000800 |
| Output | $0.80 | $0.000800 |
Live Performance
Live endpoint metrics, refreshed every 30 minutes.
Leaderboard Categories
External Resources
Frequently asked questions about DeepSeek: R1 Distill Llama 70B
How much does DeepSeek: R1 Distill Llama 70B cost?
DeepSeek: R1 Distill Llama 70B costs $0.80 per million input tokens and $0.80 per million output tokens.
What is the context window of DeepSeek: R1 Distill Llama 70B?
DeepSeek: R1 Distill Llama 70B has a context window of 128,000 tokens (128K).
Who created DeepSeek: R1 Distill Llama 70B?
DeepSeek: R1 Distill Llama 70B is developed by DeepSeek and was released on January 23, 2025.
