How much VRAM does ced-base need?

Q8_0 is exactly 132,833,376 bytes (0.12 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

mispeech · text

ced-base

mispeech/ced-base

ced-base at Q8_0 is exactly 132,833,376 bytes (0.12 GiB / 0.13 GB) — an effective 12.398 bits per weight, not the nominal 8.

From the file· summed from 4 file(s)From the file· KV per layer

Parameters

86M

Architecture

ced

null layers

Context

—

native (config.json)

License

apache-2.0

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
Q8_04 shards	0.12 GiB	132,833,376	12.398	—	mudler
F164 shards	0.23 GiB	246,980,608	23.052	—	mudler
F32	0.32 GiB	342,936,608	32.008	—	mudler

Compare with

same modality, comparable size

vs embeddinggemma-300m vs Qwen3-0.6B vs Qwen2.5-0.5B-Instruct vs SmolLM2-135M-Instruct vs gemma-3-270m-it vs Qwen3-0.6B-Base All alternatives →Will it fit on my card? →

Why other calculators give a different number

A parameters × bits ÷ 8 estimate puts Q8_0 at roughly 0.04 GiB. The real file is 0.12 GiB, because a quantization is a mixture and some tensors are always kept at higher precision.

Architecture

from config.json

Layers

—

Attention heads

—

KV heads

—

Head dim

—

Hidden size

—

Vocab

—

Sliding window

none

SWA period

—

MLA

Experts

—

Experts per token

—

use_sliding_window

—

Questions people ask

How much VRAM does ced-base need?: Q8_0 is exactly 132,833,376 bytes (0.12 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
Which quantization of ced-base should I use?: Q4_K_M is the usual default. Pick the largest quantization that fits your card at the context you actually need — the table above gives exact sizes for every one published.