DeepSeek V4 Pro (Free)

DeepSeek V4 Pro hosted free by NVIDIA. 1.6T total / 49B active MoE, 1M context, top open reasoning (MMLU-Pro 87.5, GPQA 90.1, SWE-bench 80.6)

Code Examples

from blockrun_llm import LLMClient

client = LLMClient()  # Uses BLOCKRUN_WALLET_KEY (never sent to server)
response = client.chat("nvidia/deepseek-v4-pro", "Hello!")

Pricing

Payment