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 RTX 4070 run Inkling-2-Mini?

Inkling-2-Mini·26.4B (3.5B active)·GeForce RTX 4070·12 GB VRAM

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

Yes. The RTX 4070 runs Inkling-2-Mini fully in VRAM at FP16 / BF16 (about 1.9 GB of 12 GB), at roughly 40 tokens/s.

best quant
FP16 / BF16
needs
1.9 GB
usable
12 GB
est. speed
~40 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 RTX 4070

QuantQualityMemoryFitSpeed
Q2_K Heavy loss9.7 GBfits~244 tok/s
Q3_K_M Noticeable loss14 GBoffload~82 tok/s
Q4_K_M Small loss17 GBoffload~42 tok/s
Q5_K_M Near-lossless20 GBoffload~29 tok/s
Q6_K Near-lossless23 GBoffload~22 tok/s
Q8_0 Lossless1.6 GBfits~75 tok/s
FP16 / BF16 pickLossless1.9 GBfits~40 tok/s

Speed is a bandwidth-based estimate for generating text (504 GB/s); prompt processing and long contexts are slower.

Measured speed on the RTX 4070

No community reports for this pair yet.

Report your own speed →

Run Inkling-2-Mini on the RTX 4070

Using FP16 / BF16, the best fit for RTX 4070.

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

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 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 RTX 4070 comfortably

All that fit →
ModelParamsBest quantMemorySpeedLicence
12.2BQ4_K_M≈10 GBFits in memory~38 tok/sConditions
9.7BQ8_0≈11 GBFits in memory~27 tok/sCommercial OK
13.2BQ4_K_M≈11 GBFits in memory~35 tok/sConditions
8.8BQ8_0≈10 GBFits in memory~30 tok/sCommercial OK
10.7BQ6_K≈10 GBFits in memory~32 tok/sConditions
9.4BQ8_0≈11 GBFits in memory~28 tok/sCommercial OK

Other GPUs that run Inkling-2-Mini

Frequently asked questions

Can the RTX 4070 run Inkling-2-Mini?
Yes. The RTX 4070 runs Inkling-2-Mini fully in VRAM at FP16 / BF16 (about 1.9 GB of 12 GB), at roughly 40 tokens/s.
How fast does Inkling-2-Mini run on the RTX 4070?
An estimate from memory bandwidth gives about 40 tokens/s at FP16 / BF16.
Which quantisation of Inkling-2-Mini should I use on the RTX 4070?
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.