selfhost.directory

The index of self-hostable projects — ranked, categorized and served with live updates.
Own your data and stop giving it to big tech.

philbert440 logo

Can the RTX 3090 run Qwen3.8-27B-Uncensored-Cyber?

Qwen3.8-27B-Uncensored-Cyber·27.4B·GeForce RTX 3090·24 GB VRAM

Saved in this browser, used on every model page.
Yes

Yes. The RTX 3090 runs Qwen3.8-27B-Uncensored-Cyber fully in VRAM at Q4_K_M (about 21 GB of 24 GB), at roughly 31 tokens/s.

best quant
Q4_K_M
needs
21 GB
usable
24 GB
est. speed
~31 tok/s

Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the Qwen3.8-27B-Uncensored-Cyber page →

Every quantisation of Qwen3.8-27B-Uncensored-Cyber on the RTX 3090

QuantQualityMemoryFitSpeed
Q2_K Heavy loss11 GBfits~57 tok/s
Q3_K_M Noticeable loss16 GBfits~39 tok/s
Q4_K_M pickSmall loss21 GBfits~31 tok/s
Q5_K_M Near-lossless23 GBtight~26 tok/s
Q6_K Near-lossless26 GBoffload~12 tok/s
Q8_0 Lossless30 GBoffload~4.3 tok/s
FP16 / BF16 Lossless56 GBtoo big—

Speed is a bandwidth-based estimate for generating text (936 GB/s); prompt processing and long contexts are slower.

Measured speed on the RTX 3090

No community reports for this pair yet.

Report your own speed →

Run Qwen3.8-27B-Uncensored-Cyber on the RTX 3090

Using Q4_K_M, the best fit for RTX 3090.

Ollama · Q4_K_M straight from Hugging Facedocs
ollama run hf.co/philbert440/Qwen3.8-27B-Uncensored-Cyber-GGUF:Q4_K_M
llama.cpp · OpenAI-compatible server with a web UIdocs
llama-server -hf philbert440/Qwen3.8-27B-Uncensored-Cyber-GGUF:Q4_K_M

Downloads the GGUF on first run. Installed through llama.app? The same flags work with `llama serve`.

LM Studio · Download with the lms CLIdocs
lms get philbert440/Qwen3.8-27B-Uncensored-Cyber-GGUF@Q4_K_M

Then load it from the LM Studio app or with `lms load`.

vLLM · Serve an OpenAI-compatible APIdocs
vllm serve philbert440/Qwen3.8-27B-Uncensored-Cyber
vLLM · Smaller AWQ builddocs
vllm serve philbert440/Qwen3.8-27B-Uncensored-Cyber-W4A16-AWQ
vLLM · Docker Compose (NVIDIA)docs
services:
  vllm:
    image: vllm/vllm-openai:latest
    command: philbert440/Qwen3.8-27B-Uncensored-Cyber
    ports:
      - "8000:8000"
    ipc: host
    volumes:
      - ~/.cache/huggingface:/root/.cache/huggingface
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]

API on port 8000.

SGLang · Serve an OpenAI-compatible APIdocs
sglang serve philbert440/Qwen3.8-27B-Uncensored-Cyber --host 0.0.0.0 --port 30000
SGLang · Docker Compose (NVIDIA)docs
services:
  sglang:
    image: lmsysorg/sglang:latest
    command: sglang serve philbert440/Qwen3.8-27B-Uncensored-Cyber --host 0.0.0.0 --port 30000
    ports:
      - "30000:30000"
    ipc: host
    volumes:
      - ~/.cache/huggingface:/root/.cache/huggingface
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]

API on port 30000.

Vision models that fit the RTX 3090 comfortably

All that fit →
ModelParamsBest quantMemorySpeedLicence
27.8BQ5_K_M≈22 GBFits in memory~26 tok/sCommercial OK
27.8BQ5_K_M≈22 GBFits in memory~26 tok/sCommercial OK
25.8BA4BQ5_K_M≈20 GBFits in memory~181 tok/sCommercial OK
27.4BQ4_K_M≈21 GBFits in memory~31 tok/sConditions
27.8BQ5_K_M≈22 GBFits in memory~26 tok/sCommercial OK
meta-models logoMuse-Glimmer-30Bmeta-models
29.8BQ5_K_M≈22 GBFits in memory~24 tok/sCommercial OK

Other GPUs that run Qwen3.8-27B-Uncensored-Cyber

Frequently asked questions

Can the RTX 3090 run Qwen3.8-27B-Uncensored-Cyber?
Yes. The RTX 3090 runs Qwen3.8-27B-Uncensored-Cyber fully in VRAM at Q4_K_M (about 21 GB of 24 GB), at roughly 31 tokens/s.
How fast does Qwen3.8-27B-Uncensored-Cyber run on the RTX 3090?
An estimate from memory bandwidth gives about 31 tokens/s at Q4_K_M.
Which quantisation of Qwen3.8-27B-Uncensored-Cyber should I use on the RTX 3090?
Q4_K_M: the largest one that fits. Expected quality: small loss compared with the full-precision weights.

Memory and speed are estimates from the model's size and the card's memory bandwidth; real numbers depend on the runtime, drivers and context length. Community figures are medians of user reports.