AIPAI Pentium
← Model index
audio-text-to-text

RumiLabs/rumi-cerberus-4b-mlx-4bit

RumiLabs/rumi-cerberus-4b-mlx-4bit is a audio text to text model indexed for deployment research. Estimated minimum GPU memory is 6 GB. It is publicly listed on Hugging Face.

View on Hugging Face ↗Add to comparison
Parameters4.0B
Downloads0
Likes0
Licenseapache-2.0
Librarymlx
Hardware tierConsumer
DEPLOYMENT NOTES

What to verify before deployment

  • Confirm the exact weight format and quantization.
  • Measure memory at your intended context or image size.
  • Review model-card limitations and evaluation methodology.
  • Test latency and throughput on your target runtime.
LICENSE SIGNAL

Likely commercial-friendly

License metadata is an index signal, not legal advice. Follow the repository license and any model-specific acceptable-use terms.

Verify at the source →
RUN THIS MODEL

Choose how you want to deploy

GPU CLOUD · SIMPLE START

Rent a GPU sized for this model

Start with at least 12 GB VRAM, then measure memory and latency with your workload.

Check RunPod availability ↗Referral link. AI Pentium may earn credits or commission from qualifying new users.
GPU MARKETPLACE · PRICE SHOP

Compare marketplace GPU offers

Filter Vast.ai offers for a GPU with at least 12 GB VRAM and compare availability by region.

Compare GPUs on Vast.ai ↗Referral link. AI Pentium may earn commission from qualifying new users.
ORIGINAL SOURCE

Inspect weights and model card

Verify files, license terms, limitations and usage instructions before deploying.

View on Hugging Face ↗Source link; AI Pentium does not host model weights.
SIMILAR OPTIONS

More audio-text-to-text models

audio-text-to-text

cstr/MOSS-Audio-4B-Instruct-GGUF

cstr/MOSS-Audio-4B-Instruct-GGUF is a audio text to text model indexed for deployment research. Estimated minimum GPU memory is 6 GB. It is publicly listed on Hugging Face.

↓ 1,232♡ 6
4.0B6 GB+ VRAMapache-2.0
Deployment details →
audio-text-to-textGated

snkii/Sori-1B

snkii/Sori-1B is a audio text to text model indexed for deployment research. Estimated minimum GPU memory is 6 GB. Access approval is required on Hugging Face.

↓ 296♡ 14
1.0B6 GB+ VRAMother
Deployment details →
audio-text-to-text

Zzo1312/ultravox-v0_5-llama-3_2-1b

Zzo1312/ultravox-v0_5-llama-3_2-1b is a audio text to text model indexed for deployment research. Estimated minimum GPU memory is 6 GB. It is publicly listed on Hugging Face.

↓ 3♡ 0
1.0B6 GB+ VRAMmit
Deployment details →
audio-text-to-text

RumiLabs/rumi-cerberus-4b-mlx-4bit

RumiLabs/rumi-cerberus-4b-mlx-4bit is a audio text to text model indexed for deployment research. Estimated minimum GPU memory is 6 GB. It is publicly listed on Hugging Face.

↓ 0♡ 0
4.0B6 GB+ VRAMapache-2.0
Deployment details →

Operate a GPU cloud or inference API?

Reach developers after they have selected a model and are ready to run it.

Become a platform partner