SAVRN Model Hub · Comparisons
t5-base vs t5-small
T5-base has 223M parameters and t5-small has 61M parameters; both are released under Apache License 2.0; at 16-bit, t5-base needs about 0.5 GB (1x MI300X from $1.85 an hour) and t5-small about 0.1 GB (1x MI300X from $1.85 an hour).
| Field | t5-base google-t5/t5-base | t5-small google-t5/t5-small |
|---|---|---|
| Publisher | T5 community | T5 community |
| Task | Translation | Translation |
| Modality | Text | Text |
| Parameters, as reported | 223M parameters | 61M parameters |
| Architecture | T5ForConditionalGeneration | T5ForConditionalGeneration |
| Library | transformers | transformers |
| Context length | Not stated | Not stated |
| Repository size | 4.5 GB | 2.2 GB |
| Artifact formats | safetensors, pytorch, jax, rust, tf | safetensors, onnx, pytorch, jax, rust, tf |
| License | apache-2.0 | apache-2.0 |
| Access | Open weights, no gate | Open weights, no gate |
| Memory at 16-bit (weights and margin) | 0.5 GB | 0.1 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 GB |
| Cheapest GPUs at 4-bit, per hour | 1x MI300X, $1.85 | 1x MI300X, $1.85 |
| Revision viewed | a9723ea7f1b3 | df1b051c4962 |
| Downloads reported by the hub | 1.9M | 23.8M |
| Last observed | 2026-10-13 | 2026-10-13 |
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 t5-base
Half a gigabyte of accelerator memory covers t5-base at 16-bit, with the weights at 0.4 GB, so the hardware question is not which card but how many jobs you stack on one. The cheapest setup in our data, a single MI300X with 192 GB at $1.85 per hour on-demand, sits nearly idle if this 223M-parameter translation model runs alone, so put it on a GPU already carrying other work.
Check two things first. Our record lists no context length, so test your longest translation inputs against the checkpoint directly. The 11 files total 4.5 GB across five formats; pull only the one your stack loads. Apache 2.0 permits commercial use, modification and redistribution as long as the license notices and any NOTICE file stay intact and you state significant changes, plus a patent grant from contributors. Pretraining was on C4; last update February 14, 2024.
SAVRN's Notes on t5-small
The 4-bit row reads 0.0 GB and the 16-bit row 0.1 GB, so quantizing buys nothing worth the trouble. The cheapest setup we list is a 192 GB MI300X at $1.85 an hour, which makes the only hardware question what else that card does. What you get is the 61M-parameter checkpoint of the text-to-text transfer transformer, pretrained on C4 with a mix of unsupervised and supervised tasks, labeled for translation.
Apache 2.0 means you can fine-tune it, ship the result and sell it, keeping the notices and stating what you changed. No context length is listed, so measure your longest input against the model yourself. The seven papers listed for it are task datasets, acceptability judgments, semantic textual similarity, inference, SQuAD, word in context, reading comprehension and yes-or-no questions, the trail for what the supervised side covered. The last update was 2023-06-30.
Questions
Which is larger, t5-base or t5-small?
t5-base (223M parameters) is larger than t5-small (61M parameters), by the parameter counts their publishers report.
Which is cheaper to run, t5-base or t5-small?
At 4-bit, t5-base fits on 1x MI300X from $1.85 an hour and t5-small on 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.
Can I use t5-base commercially?
Yes. t5-base 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.
Can I use t5-small commercially?
Yes. t5-small 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.