DeepSeek logo

Rent DeepSeek-R1 70B in a private loop

Flagship reasoning model
By DeepSeekdeepseek-r1:70b70B parameters16K tokens of contextMIT

Top-tier reasoning: it holds a long chain of thought where 32B already loses track and handles problems with several interacting constraints. Slow and hungry — taken for a specific class of tasks, not for general chat.

GPU configurations for DeepSeek-R1 70B

1x A100

Best value
~22 tok/sup to 3 streams
$1.48per hour

1x H100

~35 tok/sup to 6 streams
$3.33per hour
Pick a GPU
VRAM
RAM
Storage
vCPU
CPU

Billing follows server uptime by the hour.

DeepSeek-R1 70B benchmarks

as published by the model vendor
DeepSeek-R1 70B scores in public benchmarks
MATH-50094.5
AIME 202470
GPQA Diamond65.2

Other models in the private loop

FAQ

Frequently asked about renting DeepSeek-R1 70B

Which GPU does DeepSeek-R1 70B need?

The entry configuration is 1x A100, which delivers around 22 tokens per second. For more parallel requests and headroom on speed, go with 1x H100.

How much does DeepSeek-R1 70B cost?

An hour of DeepSeek-R1 70B runs from $1.48 to $3.33 depending on the GPU configuration. Billing is hourly: you pay for server time, not for tokens.

How do I call DeepSeek-R1 70B from my code?

Once deployed you get an OpenAI-compatible endpoint: point any OpenAI client at the base_url you receive and pass deepseek-r1:70b as the model name. Code written against the OpenAI API needs no changes.

Does the data stay private?

Yes. DeepSeek-R1 70B runs on a dedicated GPU server inside a private loop: requests and responses never reach the model vendor and are not used for training.

What context window and license does DeepSeek-R1 70B have?

DeepSeek-R1 70B holds 16K tokens of context per request and ships under the MIT license, so you can use it commercially on your own hosting terms.