# Can the RTX A4000 run KORMo-Web-11B?

> Yes. The RTX A4000 runs KORMo-Web-11B fully in VRAM at Q8_0 (about 13 GB of 16 GB), at roughly 21 tokens/s.

RTX A4000: 16 GB (16 GB usable), 448 GB/s. Estimates at an 8K context.

## Every quantisation

| Quant | Quality | ≈ GB | Fit | Est. tok/s |
|---|---|---|---|---|
| Q2_K | Heavy loss | 5.35 | fits | 67 |
| Q3_K_M | Noticeable loss | 7.1 | fits | 45 |
| Q4_K_M | Small loss | 8.38 | fits | 36 |
| Q5_K_M | Near-lossless | 9.53 | fits | 31 |
| Q6_K | Near-lossless | 10.76 | fits | 27 |
| Q8_0 (pick) | Lossless | 13.38 | fits | 21 |
| F16 | Lossless | 23.64 | offload | 3.6 |

## Run it

Serve an OpenAI-compatible API (vllm):

```bash
vllm serve KORMo-VL/KORMo-Web-11B
```

Docker Compose (NVIDIA) (vllm):

```yaml
services:
  vllm:
    image: vllm/vllm-openai:latest
    command: KORMo-VL/KORMo-Web-11B
    ports:
      - "8000:8000"
    ipc: host
    volumes:
      - ~/.cache/huggingface:/root/.cache/huggingface
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]
```

Ask about an image (transformers):

```python
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="KORMo-VL/KORMo-Web-11B", 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 comfortably

- [Qwen/Qwen3.5-9B](https://selfhost.directory/models/qwen/qwen3.5-9b/on/rtx-a4000): Q8_0 ≈10.81 GB, est. 24 tok/s
- [google/gemma-3-12b-it](https://selfhost.directory/models/google/gemma-3-12b-it/on/rtx-a4000): Q6_K ≈12.76 GB, est. 25 tok/s
- [Qwen/Qwen3-VL-8B-Instruct](https://selfhost.directory/models/qwen/qwen3-vl-8b-instruct/on/rtx-a4000): Q8_0 ≈10.15 GB, est. 26 tok/s
- [meta-llama/Llama-3.2-11B-Vision-Instruct](https://selfhost.directory/models/meta-llama/llama-3.2-11b-vision-instruct/on/rtx-a4000): Q8_0 ≈11.91 GB, est. 22 tok/s
- [XiaomiMiMo/MiMo-V2.6-Distill-Qwen-9B](https://selfhost.directory/models/xiaomimimo/mimo-v2.6-distill-qwen-9b/on/rtx-a4000): Q8_0 ≈10.83 GB, est. 25 tok/s
- [Qwen/Qwen2.5-VL-7B-Instruct](https://selfhost.directory/models/qwen/qwen2.5-vl-7b-instruct/on/rtx-a4000): Q8_0 ≈8.86 GB, est. 28 tok/s

## Other GPUs that run it

- [GeForce RTX 4060](https://selfhost.directory/models/kormo-vl/kormo-web-11b/on/rtx-4060): Q3_K_M, est. 27 tok/s
- [Apple M4 (16 GB)](https://selfhost.directory/models/kormo-vl/kormo-web-11b/on/apple-m4-16gb): Q5_K_M, est. 8.3 tok/s
- [Radeon RX 6700 XT](https://selfhost.directory/models/kormo-vl/kormo-web-11b/on/rx-6700-xt): Q6_K, est. 23 tok/s
- [GeForce RTX 3060 12GB](https://selfhost.directory/models/kormo-vl/kormo-web-11b/on/rtx-3060): Q6_K, est. 21 tok/s
- [Arc B580](https://selfhost.directory/models/kormo-vl/kormo-web-11b/on/arc-b580): Q6_K, est. 27 tok/s
- [Tesla T4](https://selfhost.directory/models/kormo-vl/kormo-web-11b/on/tesla-t4): Q8_0, est. 15 tok/s
- [Arc A770 16GB](https://selfhost.directory/models/kormo-vl/kormo-web-11b/on/arc-a770): Q8_0, est. 26 tok/s
- [Radeon RX 7800 XT](https://selfhost.directory/models/kormo-vl/kormo-web-11b/on/rx-7800-xt): Q8_0, est. 29 tok/s

---
Model: https://selfhost.directory/models/kormo-vl/kormo-web-11b. GPU: https://selfhost.directory/models/gpus/rtx-a4000. Source: https://selfhost.directory/models/kormo-vl/kormo-web-11b/on/rtx-a4000