How much of this pack is actually 4-bit?

A name like W4A16 or -AWQ-4bit reads as though everything in the file is four bits. It never is. Embeddings, the output head, the norms and any preserved auxiliary head stay at the source's precision — commonly a fifth to two fifths of the bytes — and no model page says which, or how much.

The set matters more than the total. An lm_head left whole is a deliberate, defensible cost. An lm_head quantized is a decision the pack made on your behalf without mentioning it: measured on our own weights, quantizing it flips the model's chosen token on about a fifth of positions.

This reads safetensors headers only, over ranged requests from your browser — a few hundred kilobytes against packs of many gigabytes. It never loads the model and never runs it.

Why this reads dtypes, not tensor names

The first version of this tool classified tensors by name, learned from compressed-tensors packs where the payload is called weight_packed. Probing nine toolchains showed that convention is the minority. bitsandbytes, ModelOpt FP8, ModelOpt NVFP4 and compressed-tensors' own int-quantized and naive-quantized formats all store the quantized payload under the plain name weight — so a name-driven reader calls every one of those packs 100% full precision, confidently, and about most of what is published.

So the dtype decides: an I32, I8, U8 or F8 tensor is not full precision whatever it is called. Names are used only to separate quantization metadata — scales, zero points, group indices — from model weights, a distinction that genuinely has no dtype signature.

The declared method is read from config.json, or from hf_quant_config.json where the tool writes it there — never from the repo name. cyankiwi/Qwen3-VL-8B-Instruct-AWQ-4bit declares compressed-tensors.

Conventions, and where each was verified

None of these naming conventions is a stable public API — they are internal choices of each tool, and they have moved before. Two versions therefore appear, answering different questions: the producer's, read out of the pack itself wherever that tool records it, which is the version that actually wrote those bytes; and this reader's, the date below, which says how stale the table is rather than leaving a wrong answer looking authoritative.