✳ OMNIVIEWER

Drop a GGUF model. Any size.

safe · secure · no server · works offline · fast

…or just start typing at the cursor.

Or try it now

↓ scroll for about & FAQ

Inspect a local LLM online — without downloading the weights

OmniViewer opens a .gguf model file — the single-file format llama.cpp, Ollama and LM Studio use — right in your browser. Drop the file and you get a plain-language architecture card: what architecture it implements, how many parameters it has, how it’s quantized and the effective bits per weight, its layer count and context length, its attention type, and — for models like Mixtral — its Mixture-of-Experts configuration. There is no upload and no server, and the heavy work runs in a background Web Worker so the page never freezes.

Why even a 40 GB model opens instantly

GGUF is header-first: everything describing the model — a key/value metadata block and a per-tensor info table — sits at the front of the file, before the raw weight data. OmniViewer reads only that bounded prefix (a few megabytes, even with a 128,000-token vocabulary) and never touches the multi-gigabyte weight body, so a huge quantized model opens as fast as a tiny one. A giant embedded vocabulary is summarized as a count instead of being loaded into memory.

What each tab does, in plain terms

For a byte-level walk through the format — the magic, the value-type enum, the tensor table and how quantization block sizes give bits-per-weight — see our deep dive into the GGUF file format.

OmniViewer opens every file format, powered by the same viewing engine as fastjsonviewer.com and hugecsv.com.

FAQ

Is my model file uploaded anywhere?

No. OmniViewer is a static page with no server-side processing: your .gguf is read directly by your browser, and every tab — architecture, metadata, tensors, tokenizer and stats — runs locally in a Web Worker. The model never leaves your computer.

How can a multi-gigabyte model open instantly?

GGUF keeps all of its metadata and its tensor table at the front of the file, before the weights. OmniViewer reads only that header prefix — a few megabytes at most — and never reads the multi-gigabyte weight body, so a 40 GB model opens as fast as a small one.

What can I learn about a model from its GGUF file alone?

A lot, with no description file: the architecture, the parameter count (computed from the tensor shapes), the quantization and effective bits-per-weight, the number of layers, the embedding and context length, the attention type (multi-head, grouped-query or multi-query), the Mixture-of-Experts configuration, the RoPE settings and the embedded tokenizer including its chat template.

Does it detect Mixture-of-Experts (MoE) models?

Yes. When the header carries an expert count (as Mixtral and other MoE models do), the architecture card shows how many experts the model has and how many are routed per token — for example, 8 experts with 2 active.

What is bits-per-weight, and how is it computed?

Bits-per-weight is the average number of bits each parameter takes on disk — a direct measure of how aggressively the model is quantized. OmniViewer computes it as the file size in bits divided by the parameter count (summed from the tensor shapes), so a Q4_K model comes out around 4.5 bits per weight.

Which quantization formats does it recognize?

All the common GGML tensor types, including F32, F16 and BF16 and the quantized types Q4_0, Q4_K, Q5_K, Q6_K, Q8_0 and the IQ family. The Tensors tab shows the exact mix, and the Architecture card names the dominant one.

Can I open a GGUF without llama.cpp or Ollama?

Yes. OmniViewer parses the GGUF structure in pure JavaScript, so it works in any modern browser on Windows, Linux, macOS or a phone. No llama.cpp, no Ollama, no Python, no install.