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.

thinkingmachines logo

Can the Radeon RX 9070 XT run Inkling-2-Mini?

Inkling-2-Mini·26.4B (3.5B active)·Radeon RX 9070 XT·16 GB VRAM

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

Yes. The Radeon RX 9070 XT runs Inkling-2-Mini fully in VRAM at FP16 / BF16 (about 1.9 GB of 16 GB), at roughly 51 tokens/s.

best quant
FP16 / BF16
needs
1.9 GB
usable
16 GB
est. speed
~51 tok/s

Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the Inkling-2-Mini page →

Every quantisation of Inkling-2-Mini on the Radeon RX 9070 XT

QuantQualityMemoryFitSpeed
Q2_K Heavy loss9.7 GBfits~310 tok/s
Q3_K_M Noticeable loss14 GBfits~208 tok/s
Q4_K_M Small loss17 GBoffload~111 tok/s
Q5_K_M Near-lossless20 GBoffload~52 tok/s
Q6_K Near-lossless23 GBoffload~33 tok/s
Q8_0 Lossless1.6 GBfits~96 tok/s
FP16 / BF16 pickLossless1.9 GBfits~51 tok/s

Speed is a bandwidth-based estimate for generating text (640 GB/s); prompt processing and long contexts are slower.

Measured speed on the Radeon RX 9070 XT

No community reports for this pair yet.

Report your own speed →

Run Inkling-2-Mini on the Radeon RX 9070 XT

Using FP16 / BF16, the best fit for Radeon RX 9070 XT.

Ollama · From the Ollama librarydocs
ollama run inkling-2-mini

The library build picks its own default quantisation.

Ollama · BF16 straight from Hugging Facedocs
ollama run hf.co/unsloth/Inkling-2-Mini-GGUF:BF16
Ollama · Docker with ROCm (AMD)docs
docker run -d --name ollama --device /dev/kfd --device /dev/dri \
  -v ollama:/root/.ollama -p 11434:11434 ollama/ollama:rocm
docker exec -it ollama ollama run inkling-2-mini

The image bundles the ROCm runtime; the host only needs the amdgpu kernel driver.

llama.cpp · OpenAI-compatible server with a web UIdocs
llama-server -hf unsloth/Inkling-2-Mini-GGUF:BF16

Needs a HIP (ROCm) build of llama.cpp. Downloads the GGUF on first run.

llama.cpp · Docker with ROCm (AMD)docs
docker run -d --name llama --device /dev/kfd --device /dev/dri -p 8080:8080 \
  -v ~/.cache/llama.cpp:/root/.cache/llama.cpp \
  ghcr.io/ggml-org/llama.cpp:server-rocm \
  -hf unsloth/Inkling-2-Mini-GGUF:BF16 --host 0.0.0.0 --port 8080 -ngl 99

Server and web UI on port 8080.

LM Studio · Download with the lms CLIdocs
lms get unsloth/Inkling-2-Mini-GGUF@BF16

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

Transformers · Ask about an imagedocs
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="thinkingmachines/Inkling-2-Mini", 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 Radeon RX 9070 XT comfortably

All that fit →
ModelParamsBest quantMemorySpeedLicence
12.2BQ6_K≈13 GBFits in memory~35 tok/sConditions
19.4BA3BQ4_K_M≈13 GBFits in memory~194 tok/sCommercial OK
9.7BQ8_0≈11 GBFits in memory~34 tok/sCommercial OK
13.2BQ6_K≈13 GBFits in memory~32 tok/sConditions
8.8BQ8_0≈10 GBFits in memory~38 tok/sCommercial OK
10.7BQ8_0≈12 GBFits in memory~31 tok/sConditions

Other GPUs that run Inkling-2-Mini

Frequently asked questions

Can the Radeon RX 9070 XT run Inkling-2-Mini?
Yes. The Radeon RX 9070 XT runs Inkling-2-Mini fully in VRAM at FP16 / BF16 (about 1.9 GB of 16 GB), at roughly 51 tokens/s.
How fast does Inkling-2-Mini run on the Radeon RX 9070 XT?
An estimate from memory bandwidth gives about 51 tokens/s at FP16 / BF16.
Which quantisation of Inkling-2-Mini should I use on the Radeon RX 9070 XT?
FP16 / BF16: the largest one that fits. Expected quality: lossless 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.