Loading…
Loading…
Models / whisper-large / whisper-large ~1.54B
Speech to text · served with vLLM · ↓ 60.5K downloads a month · View on Hugging Face
RedHatAI/whisper-large-v3-FP8-dynamic needs about 4 GB of GPU memory. It runs on one RTX 3060 from $0.09/h right now; best value: RTX 5070 Ti at $0.19/h.
The same model packaged differently: a quantized version needs less GPU memory and runs on a cheaper machine.
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|---|---|---|---|
| Originalopenai/whisper-large-v3 | vLLM | 9.3 GB | 13.1 GB | $0.13/h RTX A4000 |
| turboopenai/whisper-large-v3-turbo | vLLM | 1.6 GB | 3.9 GB | $0.09/h RTX 3060 |
| FP8 dynamic (RedHatAI)RedHatAI/whisper-large-v3-FP8-dynamic | vLLM | 1.7 GB | 4 GB | $0.09/h RTX 3060 |
| kb whisper large (KBLab)KBLab/kb-whisper-large | vLLM | 3.2 GB | 5.9 GB | $0.09/h RTX 3060 |
| turbo (onnx-community)onnx-community/whisper-large-v3-turbo | ONNX Runtime | 190 MB | 1.3 GB | $0.07/h Quadro P4000 |
| turbo timestamped (onnx-community)onnx-community/whisper-large-v3-turbo_timestamped | ONNX Runtime | 190 MB | 1.3 GB |
$0.19/h from your balance while it runs, billed by the hour; stop it from its page. GridPup rents the machine, which downloads the model from Hugging Face (usually ready in 10 to 15 minutes), and gives you an https address, keys and a Playground.
| $0.07/h Quadro P4000 |
| turbo gguf (xkeyC)xkeyC/whisper-large-v3-turbo-gguf | ONNX Runtime | 1.8 GB | 3.6 GB | $0.07/h Quadro P4000 |
| turbo (TheStageAI)TheStageAI/thewhisper-large-v3-turbo | vLLM | 1.6 GB | 3.9 GB | $0.09/h RTX 3060 |
Prices are per hour for the whole machine and its disk, right now. Speed is Vast's DLPerf score of the machine, relative to the cheapest. Best value: the most speed per dollar. Only machines with a Vast reliability above 95% that download at 400 Mbit/s or more are offered: the machine fetches the model itself.