SAVRN Model Hub · Comparisons
bert-base-arabertv02 vs bert-base-multilingual-uncased
Bert-base-arabertv02 has 136M parameters and bert-base-multilingual-uncased has 168M parameters; at 16-bit, bert-base-arabertv02 needs about 0.3 GB (1x MI300X from $1.85 an hour) and bert-base-multilingual-uncased about 0.4 GB (1x MI300X from $1.85 an hour).
| Field | bert-base-arabertv02 aubmindlab/bert-base-arabertv02 | bert-base-multilingual-uncased google-bert/bert-base-multilingual-uncased |
|---|---|---|
| Publisher | AUB MIND LAB | BERT community |
| Task | Fill mask | Fill mask |
| Modality | Text | Text |
| Parameters, as reported | 136M parameters | 168M parameters |
| Architecture | BertForMaskedLM | BertForMaskedLM |
| Library | transformers | transformers |
| Context length | 512 tokens | 512 tokens |
| Repository size | 3.0 GB | 3.0 GB |
| Artifact formats | safetensors, pytorch, tensorboard, jax, tf | safetensors, pytorch, jax, tf |
| License | Not stated | apache-2.0 |
| Access | Open weights, no gate | Open weights, no gate |
| Memory at 16-bit (weights and margin) | 0.3 GB | 0.4 GB |
| Cheapest GPUs at 16-bit, per hour | 1x MI300X, $1.85 | 1x MI300X, $1.85 |
| Memory at 4-bit (weights and margin) | 0.1 GB | 0.1 GB |
| Cheapest GPUs at 4-bit, per hour | 1x MI300X, $1.85 | 1x MI300X, $1.85 |
| Revision viewed | 016fb9d6768f | 7cbf9a625e29 |
| Downloads reported by the hub | 567.2k | 2.1M |
| Last observed | 2026-09-18 | 2026-10-14 |
An evaluation row appears only where at least two of these models report the same benchmark with the same stated configuration, metric, unit and setup. Different evaluators stay named in each cell. Values are shown as reported: no unit conversion, no ranking.
SAVRN's Notes on bert-base-arabertv02
If the text is Arabic and the job is masked-token prediction or a base for fine-tuning, AUB MIND LAB's 136M-parameter BERT-base model, trained on Wikipedia, Osian, the 1.5B-Arabic-Corpus, oscar-arabic-unshuffled and a private Assafir set, is the one on this page. Memory is a rounding error: 0.3 GB at 16-bit, 0.2 GB at 8-bit, 0.1 GB at 4-bit. The Index's cheapest host is a 192 GB MI300X at $1.85 per hour; you would never dedicate one to this.
Here is the item that decides the deployment: the facts we hold list no license for this model. Open access means the files can be fetched, but with no stated terms we treat commercial use, modification and redistribution as unresolved until the publisher confirms them. Check that first. Then note the 512-token context, which means chunking anything longer than a page, and 24 files across five formats, so pick one.
SAVRN's Notes on bert-base-multilingual-uncased
Four tenths of a gigabyte in 16-bit. For a 168M-parameter fill-mask model that number settles the hardware question: the cheapest setup we list, one MI300X with 192 GB at $1.85 an hour, would sit nearly idle, so put it on a card you already own as a tenant, or run 8-bit at 0.2 GB. It reads 512 tokens per pass, was pretrained on the 102 largest Wikipedia languages, and is uncased, so english and English look the same to it.
Apache 2.0 clears commercial use and redistribution provided the license, copyright notices and any NOTICE file stay attached and significant changes are stated. The 3.0 GB download spans 10 files in four formats, safetensors, pytorch, jax and tf, and you need only one. The card was written by someone other than the team that released the weights, so the paper, arXiv:1810.04805, is the source to read.
Questions
Which is larger, bert-base-arabertv02 or bert-base-multilingual-uncased?
bert-base-multilingual-uncased (168M parameters) is larger than bert-base-arabertv02 (136M parameters), by the parameter counts their publishers report.
Which is cheaper to run, bert-base-arabertv02 or bert-base-multilingual-uncased?
At 4-bit, bert-base-arabertv02 fits on 1x MI300X from $1.85 an hour and bert-base-multilingual-uncased on 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.
Can I use bert-base-multilingual-uncased commercially?
Yes. bert-base-multilingual-uncased is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.