BTC
ETH
USDT
XMR
LTC
Open model · Qwen · Apache 2.0
Run Qwen3.8 27B
on your own GPUs.
Latest dense Qwen: vision, adjustable thinking, agentic coding. 27B parameters, 56 GB of official BF16 weights.
- 67 GBGPU memory, BF16
- 34 GBGPU memory, FP8
- 1× RTX 4090cheapest server, Ollama
- $269per month
How much VRAM Qwen3.8 27B needs.
Weights × 1.2 for the KV cache and activations, with 92% of GPU memory usable: the rule our deploy page uses. Long contexts and many parallel requests need more.
| Serving option | GPU memory | Cheapest server | Per month | Action |
|---|---|---|---|---|
| vLLM or SGLang, official BF16 weights | 67 GB | 1× NVIDIA RTX PRO 6000 | $959/mo | Deploy |
| vLLM or SGLang, FP8 | 34 GB | 2× NVIDIA RTX 4090 | $538/mo | Deploy |
| Ollama + Open WebUI, qwen3.8:27b (quantized) | 22 GB | 1× NVIDIA RTX 4090 | $269/mo | Deploy |
Weights: Qwen/Qwen3.8-27B · 56 GB, BF16 · Ollama: qwen3.8:27b, 18 GB. How we estimate GPU memory
Which GPUs run Qwen3.8 27B.
Smallest server of each GPU that holds the model, and its monthly price.
| GPU | BF16 | FP8 | Ollama | Price, BF16 |
|---|---|---|---|---|
| NVIDIA B300288 GB HBM3e | 1× | 1× | 1× | $4,019/mo |
| NVIDIA B200180 GB HBM3e | 1× | 1× | 1× | $3,439/mo |
| AMD MI355X288 GB HBM3E | n/a | n/a | 1× | — |
| NVIDIA H200141 GB HBM3e | 1× | 1× | 1× | $2,279/mo |
| NVIDIA H10080 GB HBM3 | 1× | 1× | 1× | $1,779/mo |
| NVIDIA RTX PRO 600096 GB GDDR7 ECC | 1× | 1× | 1× | $959/mo |
| NVIDIA L40S48 GB GDDR6 ECC | 2× | 1× | 1× | $1,578/mo |
| NVIDIA RTX 509032 GB GDDR7 | 4× | 2× | 1× | $1,316/mo |
| NVIDIA RTX 409024 GB GDDR6X | 4× | 2× | 1× | $1,076/mo |
| NVIDIA RTX 508016 GB GDDR7 | — | 4× | 2× | — |
n/a = ROCm support for this model is not confirmed with vLLM or SGLang. — = larger than the biggest server of that GPU. Size another model
Deploy Qwen3.8 27B preinstalled.
Pick the model at the Software step of the deploy page: we install it with the engine you choose, at no extra cost.
- 1
Pick the server
The deploy page proposes the cheapest server that holds the model, and switches when you change the precision or the engine.
- 2
Pick the engine
vLLM or SGLang serve an OpenAI-compatible API; Ollama with Open WebUI gives a private chat in the browser.
- 3
Pay and log in
Pay the month in BTC, ETH, USDT, XMR or LTC, no KYC. Your server is online in under 10 minutes after confirmation, with root SSH access.
Guides: serving LLMs with vLLM · GPU servers for LLM inference · first steps on your server
Qwen3.8 27B FAQ.
Every model: open models and their VRAM.
How much VRAM does Qwen3.8 27B need?
About 67 GB of GPU memory to serve the official BF16 weights (56 GB) with vLLM or SGLang, by our rule of weights × 1.2 for the KV cache and activations; about 34 GB in FP8; about 22 GB with Ollama’s qwen3.8:27b build. Long contexts and many parallel requests need more.
Can Qwen3.8 27B run on a single GPU?
Yes, on one B300, B200, MI355X, H200, H100, RTX PRO 6000, L40S, RTX 5090 or RTX 4090 with Ollama + Open WebUI, qwen3.8:27b (quantized), by our sizing rule.
What is the cheapest server for Qwen3.8 27B?
The cheapest CryptGPU server for Qwen3.8 27B is 1× NVIDIA RTX 4090 at $269 a month, with Ollama + Open WebUI, qwen3.8:27b (quantized). The model can be preinstalled at no extra cost when you order, and you pay in crypto with no KYC.
Does Qwen3.8 27B run on the AMD MI355X?
Not confirmed with vLLM or SGLang on ROCm yet, so we offer it on NVIDIA servers; with Ollama it also runs on the MI355X.
What licence does Qwen3.8 27B use?
Apache 2.0. We download the official weights from Hugging Face (Qwen/Qwen3.8-27B); you are responsible for the licence.
Your own Qwen3.8 27B, ready in minutes.
Dedicated GPUs, one monthly price, the model preinstalled. Paid in crypto, no KYC.