Skip to main content

Model requirements

What a model needs to run is mostly decided before you fetch it: the variant and quantization fix the weight size, and the context length fixes the working memory on top. This page carries the figures for the models the documentation uses and the classes of hardware they fit; clika-modelverse info <source> --dry gives the on-disk number for any source, and the bench command measures the rest on your machine.

How to read the tables: Disk is the snapshot size (fetch --dry total). Run is the resident memory serving one session at the default context; longer contexts and concurrent sessions add KV cache on top, roughly linear in tokens. Fits names the smallest sensible device class: phone (4 GB), laptop or small board (8 GB), workstation GPU (8 GB VRAM and up).

Text generation

ModelVariantDiskRunFits
Llama 3.2 1B Instructfp16 (safetensors)2.3 GB3.0 GBlaptop
Llama 3.2 1B InstructGGUF Q6_K1.0 GB1.6 GBphone
Llama 3.1 8B InstructGGUF Q4_K_M4.9 GB6.1 GBworkstation GPU, 8 GB
Qwen3 0.6Bfp16 (safetensors)1.4 GB1.9 GBphone
Qwen3 4BGGUF Q6_K3.3 GB4.3 GBlaptop
Qwen3 4BGGUF Q8_04.3 GB5.4 GBworkstation GPU, 8 GB

The quantization ladder trades memory for fidelity in predictable steps: Q8_0 is near-fp16, Q6_K is the everyday default, Q4_K_M is the last stop before quality degrades noticeably in chat use. --kv-quant shrinks the per-token cache the same way when long contexts dominate the budget.

Multimodal

ModelVariantDiskRunFits
Gemma 3 4B IT (vision)fp16 (safetensors)8.6 GB10.2 GBworkstation GPU, 12 GB
Gemma 3 4B IT (vision)GGUF Q6_K3.9 GB5.6 GBlaptop

A multimodal run holds the vision tower alongside the language model; the difference between the text-only 4B rows above and these is that tower.

Speech-to-text

ModelVariantDiskRunFits
Whisper basefp16 (safetensors)0.3 GB0.6 GBphone
Whisper large-v3-turbofp16 (safetensors)1.6 GB2.2 GBlaptop
Whisper large-v3fp16 (safetensors)3.1 GB4.0 GBlaptop

Transcription memory is stable with audio length (the model works in 30-second windows); throughput, not memory, is what a faster device buys.

Embeddings

ModelVariantDiskRunFits
Qwen3 Embedding 0.6BGGUF Q8_00.7 GB1.1 GBphone

Embedding models run one forward per input with no generation loop, so batch size is the only memory knob that matters in practice.

Reading a figure you do not see here

For any other model: info --dry prints the exact disk cost before anything moves, devices prints what this machine offers, and a Run estimate is the disk figure plus 25 to 40 percent for activations and the default context. When the estimate is close to the device's limit, measure with bench before committing the deployment; it reports peak resident memory along with throughput.