Model reference · open weights

aya-expanse

LLMs CohereLabs Text gen 2 builds Non-commercial 15k dl/mo

aya-expanse is an open-weight language model from Cohere. aya-expanse-32b (FP16) weighs 64.6 GB; the smallest configuration that runs it is H100 80 GB.

Request Rent a GPU

For self-hosting, compare this build's requirements below with the GPU you plan to rent.

Aya Expanse 8B is an open-weight text generation model developed by Cohere Labs with 8.0 billion parameters.
It is designed for multilingual tasks and supports 23 languages, including English, French, German, Spanish, Italian, Portuguese, Japanese, and Korean.
The model has a context length of 8K and is released under the cc-by-nc-4.0 license.

Summary of the CohereLabs/aya-expanse-8b model card, 2026-10-01. The estimate below is for another build of the family.

Hugging Face ↗GitHub ↗Paper ↗

What it is

Released by	Cohere
Released	2024-10-23
Parameters	8.0B
VRAM	64.6 GB for the weights

What it runs on

Memory and cards for aya-expanse-32b (FP16)

64.6 GBweights, file size

762 MBruntime overhead, at least

How much memory each request adds isn't estimated yet for this architecture. The weights need at least the cards below, plus room for the context.

Card	Weights alone
RTX 3060 12 GB … L40S 48 GB 6 smaller cards	does not fit
A100 80 GB	fits
H100 80 GB	tight
RTX PRO 6000 Blackwell 96 GB	fits
DGX Spark (GB10) 128 GB unified	fits
H200 141 GB	fits
B200 180 GB	fits

Builds

Sizes, precisions & builds

Build	Params	Precision	Weights	Smallest setup
aya-expanse-8b ↗	8.0B	FP16	16.1 GB	RTX 4090 24 GB
aya-expanse-32b (above) ↗	32.3B	FP16	64.6 GB	H100 80 GB

How it works

How language models work

Details

Published under	CohereLabs
Type	Language models
Task	Text gen
Runs with	transformers
Popularity	15k downloads / month
Weights	64.6 GB, aya-expanse-32b (FP16, file size)

Languages

enfrdeesitptjakozharelfaplid