llamaperf

DeepSeek V2

2 reports

DeepSeek V2 VRAM requirements by size and quant →

How does DeepSeek V2 run on your hardware?

Pick your GPU or Mac, then read the speeds people reported for DeepSeek V2 on it, or estimate memory fit and speed.

Free to use. No account needed. Memory estimates and community measurements are labelled separately.

Run DeepSeek V2 yourself? Add your numbers.

Your coding agent can run a speed test on your machine and send it here with a free account. Or paste a result you've already posted.

Reported performance for DeepSeek V2

Filter this model’s reports by setup →
2 reports for this model so far, so treat the figures as early. Add yours.

Speeds are as each source reported them. GPU count, offloading and concurrent requests change them, and prompt-processing speed is input throughput, not time to first token. Check the full setup before comparing two reports.

DeepSeek V2 Lite-Chat

Unknown GPU · Project Zero

Tone: positive
generation:
11.7-12.6 tokens/s
quant:
Q4_K_S (GGUF)

User reports Project Zero v0.2.0 running DeepSeek-V2-Lite-Chat Q4_K_S at 11.7-12.6 tok/s on a 4-vCPU AVX-512 VNNI Xeon at four threads, up from 6.3-7.3 tok/s. The engine is CPU-only, written in C with no ML framework dependency; the same host recorded llama.cpp at 9.4-10.5 tok/s, and resident memory fell from 12.0 GB to 10.6 GB. With four concurrent clients, aggregate throughput rose from 45 to 101 tok/s for SmolLM2-135M Q8_0 and from 12.7 to 29.8 tok/s for SmolLM2-1.7B Q4_K_M. Startup time dropped to roughly 1-8 seconds.

Oct 24, 2026

DeepSeek V2 16B (2.4B active) Lite

Unknown GPU · llama.cpp

Tone: mixed
generation:
13.8 tokens/s
quant:
Q4_K_S

User reports DeepSeek-V2-Lite-Chat Q4_K_S at 13.79 tok/s with llama.cpp on an Intel Core i5-11300H CPU at 4 threads. The user's own C99 engine runs the same model at 1.90 tok/s on the same hardware, and the post is about closing that gap.

Oct 8, 2026

Get a weekly email of new DeepSeek V2 reports on any GPU.

Email me new reports