Unverified50% confidenceBenchmarkExact time
Stratum runs the DeepSeek 2.0 BPW version on a configuration with a 12G 5070 graphics card and 64G of memory, outputting about 6 to 4 tokens per second.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/31/2026
First Seen
Valid until: 11/14/2026
Sources
Related Entities
Related Claims
UnverifiedWith a 5070 12 GB GPU paired with 64 GB of memory, speed can reach 93 tokens/s in short chat scenarios and 74 tokens/s for 128K long contexts.73% similarUnverifiedIn the DeepSeek V1 era, each incremental token required nearly 400 KB of KV cache, which dropped to 48 KB in V3.2.72% similarUnverifiedThis 1.75-bit quantized DeepSeek model can run locally on a consumer-grade machine with a single 12GB-VRAM graphics card plus system memory.71% similarUnverifiedStrata 0.2.4 exploits MoE sparsity to run DeepSeek V4.1 Flash on a 12GB graphics card by placing infrequently used expert weights in memory or on SSD and keeping only high-frequency experts resident in VRAM.69% similarUnverifiedDeveloper Soto wrote his own inference engine and ran the 552B-parameter DeepSeek V4.1 Flash on a gaming PC with a single 24G RTX 4090, 90G of memory plus NVMe storage, measuring about 27 tokens/s decoding and 180-320 tokens/s prefill.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1120201API
curl https://kongchang.com/api/v1/knowledge/claims/1120201MCP
get_claim(id=1120201)