Open-weight model
voxcpm2-multilang28-full
by Waxal Benchmarking waxal-benchmarking/voxcpm2-multilang28-full
voxcpm2-multilang28-full is an open-weight model from Waxal Benchmarking. It has 2.3B parameters. At 16-bit it needs about 5.5 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index. It draws 27 downloads a month.
Runs On
What it takes to serve voxcpm2-multilang28-full (2.3B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.
| Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits |
|---|---|---|---|---|---|
| 16-bit | 4.6 GB | 5.5 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 2.3 GB | 2.7 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 1.1 GB | 1.4 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Oct 18, 2026.
voxcpm2-multilang28-full on every accelerator the SAVRN Index prices, at every precision
Model Card
The publisher has not written a card for this model.
Identity and Version
- Repository
- waxal-benchmarking/voxcpm2-multilang28-full
- Publisher
- Waxal Benchmarking
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- 2.3B parameters
- Languages
- Not stated by the source
- Revision
- f806ad492e03577b596443ee74551cdb102cf739
- First published
- 2026-10-12
- Last updated
- 2026-10-13
Files and Weights
10 files, 27.9 GB in total. The weights are 4 files totalling 27.9 GB in pth, safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| audiovae.pth | Weights | 377.0 MB | 94b5d51e107e |
| model.safetensors | Weights | 9.2 GB | 76146add313f |
| optimizer.pth | Weights | 18.3 GB | 71f918422967 |
| scheduler.pth | Weights | 1.5 KB | 7756f52add97 |
| config.json | Configuration | 6.6 KB | — |
| special_tokens_map.json | Configuration | 1.6 KB | — |
| training_state.json | Configuration | 15 B | — |
| .gitattributes | Repository | 1.5 KB | — |
| tokenizer.json | Tokenizer | 6.7 MB | — |
| tokenizer_config.json | Tokenizer | 1.1 KB | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 27.9 GB
Released by Waxal Benchmarking through its official repository on Hugging Face.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 27.9 GB |
| 16-bit | 4.6 GB |
| 8-bit | 2.3 GB |
| 4-bit | 1.1 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About voxcpm2-multilang28-full
How much GPU memory does voxcpm2-multilang28-full need?
About 5.5 GB at 16-bit and 1.4 GB at 4-bit: the weights (2.3B parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run voxcpm2-multilang28-full on?
At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.