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.

h2oai logo

Can the Radeon RX 7900 XTX run h2o-lightning-31b?

h2o-lightning-31b·31.3B·Radeon RX 7900 XTX·24 GB VRAM

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

Yes. The Radeon RX 7900 XTX runs h2o-lightning-31b fully in VRAM at Q4_K_M (about 21 GB of 24 GB), at roughly 28 tokens/s.

best quant
Q4_K_M
needs
21 GB
usable
24 GB
est. speed
~28 tok/s

Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the h2o-lightning-31b page →

Every quantisation of h2o-lightning-31b on the Radeon RX 7900 XTX

QuantQualityMemoryFitSpeed
Q2_K Heavy loss13 GBfits~51 tok/s
Q3_K_M Noticeable loss17 GBfits~35 tok/s
Q4_K_M pickSmall loss21 GBfits~28 tok/s
Q5_K_M Near-lossless24 GBoffload~20 tok/s
Q6_K Near-lossless28 GBoffload~6.8 tok/s
Q8_0 Lossless35 GBoffload~2.8 tok/s
FP16 / BF16 Lossless64 GBtoo big—

Speed is a bandwidth-based estimate for generating text (960 GB/s); prompt processing and long contexts are slower.

Measured speed on the Radeon RX 7900 XTX

No community reports for this pair yet.

Report your own speed →

Run h2o-lightning-31b on the Radeon RX 7900 XTX

vLLM · Serve an OpenAI-compatible APIdocs
vllm serve h2oai/h2o-lightning-31b

Gated on Hugging Face: accept the licence on the model page and export HF_TOKEN first. Needs vLLM built for ROCm; it supports Instinct MI200/MI300 and Radeon RX 7900 / 9000 cards.

vLLM · Docker Compose (AMD ROCm)docs
services:
  vllm:
    image: rocm/vllm:latest
    command: vllm serve h2oai/h2o-lightning-31b --host 0.0.0.0 --port 8000
    ports:
      - "8000:8000"
    ipc: host
    devices:
      - /dev/kfd
      - /dev/dri
    group_add:
      - video
    security_opt:
      - seccomp=unconfined
    volumes:
      - ~/.cache/huggingface:/root/.cache/huggingface
    environment:
      - HF_TOKEN=${HF_TOKEN}

AMD’s rocm/vllm image. API on port 8000.

SGLang · Serve an OpenAI-compatible APIdocs
sglang serve h2oai/h2o-lightning-31b --host 0.0.0.0 --port 30000

Gated on Hugging Face: accept the licence on the model page and export HF_TOKEN first. Needs SGLang’s ROCm build (Instinct MI300-class GPUs).

Transformers · Ask about an imagedocs
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="h2oai/h2o-lightning-31b", 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"])

Gated on Hugging Face: accept the licence on the model page and export HF_TOKEN first.

Vision models that fit the Radeon RX 7900 XTX comfortably

All that fit →
ModelParamsBest quantMemorySpeedLicence
27.8BQ5_K_M≈22 GBFits in memory~27 tok/sCommercial OK
36BA3BQ4_K_M≈22 GBFits in memory~290 tok/sCommercial OK
meta-models logoMuse-Glimmer-30Bmeta-models
29.8BQ5_K_M≈22 GBFits in memory~25 tok/sCommercial OK
27.8BQ5_K_M≈22 GBFits in memory~27 tok/sCommercial OK
25.8BA4BQ5_K_M≈20 GBFits in memory~186 tok/sCommercial OK
36BA3BQ4_K_M≈23 GBFits in memory~290 tok/sCommercial OK

Other GPUs that run h2o-lightning-31b

Frequently asked questions

Can the Radeon RX 7900 XTX run h2o-lightning-31b?
Yes. The Radeon RX 7900 XTX runs h2o-lightning-31b fully in VRAM at Q4_K_M (about 21 GB of 24 GB), at roughly 28 tokens/s.
How fast does h2o-lightning-31b run on the Radeon RX 7900 XTX?
An estimate from memory bandwidth gives about 28 tokens/s at Q4_K_M.
Which quantisation of h2o-lightning-31b should I use on the Radeon RX 7900 XTX?
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.