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.

inclusionAI logo

Can the RTX 4060 run ZwZ2-9B?

ZwZ2-9B·9.4B·GeForce RTX 4060·8 GB VRAM

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

Yes. The RTX 4060 runs ZwZ2-9B fully in VRAM at Q4_K_M (about 7.0 GB of 8.0 GB), at roughly 26 tokens/s.

best quant
Q4_K_M
needs
7.0 GB
usable
8.0 GB
est. speed
~26 tok/s

Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the ZwZ2-9B page →

Every quantisation of ZwZ2-9B on the RTX 4060

QuantQualityMemoryFitSpeed
Q2_K Heavy loss4.5 GBfits~48 tok/s
Q3_K_M Noticeable loss6.0 GBfits~33 tok/s
Q4_K_M pickSmall loss7.0 GBfits~26 tok/s
Q5_K_M Near-lossless7.8 GBtight~22 tok/s
Q6_K Near-lossless8.7 GBoffload~15 tok/s
Q8_0 Lossless11 GBoffload~7.8 tok/s
FP16 / BF16 Lossless19 GBoffload~2.6 tok/s

Speed is a bandwidth-based estimate for generating text (272 GB/s); prompt processing and long contexts are slower.

Measured speed on the RTX 4060

No community reports for this pair yet.

Report your own speed →

Run ZwZ2-9B on the RTX 4060

Using Q4_K_M, the best fit for RTX 4060.

llama.cpp · OpenAI-compatible server with a web UIdocs
llama-server -hf inclusionAI/ZwZ2-9B-GGUF:Q4_K_M

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

vLLM · Serve an OpenAI-compatible APIdocs
vllm serve inclusionAI/ZwZ2-9B
vLLM · Docker Compose (NVIDIA)docs
services:
  vllm:
    image: vllm/vllm-openai:latest
    command: inclusionAI/ZwZ2-9B
    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 inclusionAI/ZwZ2-9B --host 0.0.0.0 --port 30000
SGLang · Docker Compose (NVIDIA)docs
services:
  sglang:
    image: lmsysorg/sglang:latest
    command: sglang serve inclusionAI/ZwZ2-9B --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.

Transformers · Ask about an imagedocs
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="inclusionAI/ZwZ2-9B", dtype="auto", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/bee.jpg"},
            {"type": "text", "text": "Describe this image."},
        ],
    }
]
out = pipe(text=messages, max_new_tokens=256)
print(out[0]["generated_text"][-1]["content"])

Vision models that fit the RTX 4060 comfortably

All that fit →
ModelParamsBest quantMemorySpeedLicence
9.7BQ4_K_M≈7.0 GBFits in memory~26 tok/sCommercial OK
8.8BQ5_K_M≈7.3 GBFits in memory~24 tok/sCommercial OK
9.4BQ4_K_M≈7.2 GBFits in memory~26 tok/sCommercial OK
8.3BQ6_K≈7.0 GBFits in memory~22 tok/sCommercial OK
Cloudflare logoclef-flashCloudflare
9.4BQ4_K_M≈7.4 GBFits in memory~26 tok/sCommercial OK
8.7BQ5_K_M≈7.3 GBFits in memory~24 tok/sCommercial OK

Other GPUs that run ZwZ2-9B

Frequently asked questions

Can the RTX 4060 run ZwZ2-9B?
Yes. The RTX 4060 runs ZwZ2-9B fully in VRAM at Q4_K_M (about 7.0 GB of 8.0 GB), at roughly 26 tokens/s.
How fast does ZwZ2-9B run on the RTX 4060?
An estimate from memory bandwidth gives about 26 tokens/s at Q4_K_M.
Which quantisation of ZwZ2-9B should I use on the RTX 4060?
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.