Unverified50% confidenceBenchmarkExact time
Running on 4 DGX Spark units with VLLM's Tensor Parallel 4, measured decoding speeds were approximately 70 tokens/second for code and HTML, approximately 45 for JSON, and approximately 30 for prose and planning tasks, while DeepSeek's own API can exceed 300 tokens/second.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/15/2026
First Seen
Valid until: 12/14/2026
Sources
Related Entities
Related Claims
UnverifiedUsing the DeepSeek Flash API can achieve a speed of approximately 300 tokens per second.69% similarVerifiedDeepSeek V4 Flash has a rate limit of 10 requests per minute (10 RPM) and 100K tokens per minute (100K TPM) on the OpenModel platform.68% similarUnverified使用Gemma 4 2B版本本地运行Codex时，处理速度大约每秒30到40个Token67% similarUnverifiedOn high-performance laptops, quantized Gemma models can achieve token generation speeds of 20–50 tokens per second; pure CPU inference is approximately 5–15 tokens/second, and INT4 quantization on NVIDIA RTX 4060/4070 can boost inference speed by 3–5 times.66% similarUnverified在配备96GB显存的M2 Max上以1-bit精度运行时，简单任务生成速度稳定在约28-30 tokens/秒66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/923769API
curl https://kongchang.com/api/v1/knowledge/claims/923769MCP
get_claim(id=923769)